Event owner lens
Use this record to benchmark whether a comparable event makes the work after the room visible. The score is 38/100, so the next move is to benchmark the weakest pillars before repeating the format.
lead or directory source agenda analysis
2026 Conference
This lead or directory source in Technology / AI / Startup shows 80 visible agenda rows from iclr.cc and scores 38/100: a thin but inspectable design signal. The clearest public signals sit in Future-of-Work Fit and Learning Transfer; the main limits are Follow Through and Participation Architecture. Visible mechanisms include Participant work, Feedback, Network design, and Learning transfer. The public record does not show follow-up or tracking, so the score should be read as design intent rather than durable... A practical reading: For a reader, this is mainly a warning or source-evidence record: it reads as a lead or directory source, with the strongest visible signal in future-of-work fit and learning transfer and the biggest open question around follow through and participation architecture. The practical test is whether the published agenda connects the room to post-event continuation and evidence. This page is an original public-evidence analysis, not a copy of the source agenda or an endorsement of the event. The score places the visible agenda in the thin outcome architecture band. The strongest visible pillars are Future-of-Work Fit, Learning Transfer, and Evidence Maturity; the thinnest visible pillars are Follow Through, Participation Architecture, and Network Design. Visible mechanisms include Participant work, Feedback, Network design, Learning transfer, and Personalization. The extracted agenda preview includes 140 visible rows. The most common formats are Presentation, Unknown, and Poster Session; the most common inferred purposes are Knowledge Transfer, Unknown, and Showcase.
Eight-pillar fingerprint
Hover any pillar to see what it measures and, where it scored low, what the agenda is missing.
Participation Architecture?30
Participation Architecture - 30/100. Participant work, contribution, interaction, and alternatives to passive broadcast.Missing: Turn passive airtime into participant work: practice, sensemaking, decisions, critique, or artifact creation.
Follow Through?5
Follow Through - 5/100. Owners, dates, commitments, progress checks, and accountability after the room.Missing: Add named owners, dates, implementation checkpoints, and a visible post-event continuation path.
Problem Specificity?43
Problem Specificity - 43/100. A clear costly problem, objective, decision, or performance target.
Personalization?44
Personalization - 44/100. Role, path, goal, preparation, or connection tailoring for participants.
Network Design?34
Network Design - 34/100. Structured weak ties, bridge-building, mixers, and relationship persistence.Missing: Replace generic networking blocks with designed introductions, ask-offer exchanges, peer groups, or bridge-building rituals.
Learning Transfer?48
Learning Transfer - 48/100. Applied practice, feedback, workplace use, refreshers, and 30-90 day transfer.
Evidence Maturity?45
Evidence Maturity - 45/100. Baseline, comparison, follow-up, isolation, and attribution confidence.
Future-of-Work Fit?61
Future-of-Work Fit - 61/100. Value against time, hybrid reality, accessibility, AI, and meeting load.
Fix the gaps
Field-tested exercises matched to this agenda's weakest pillars, from the exercise library.
Follow Through (5/100)
Exercises that strengthen it: Mental Toughness Workshop • WorkshopBank · 15% Solutions • WorkshopBank · Network Patches
Participation Architecture (30/100)
Exercises that strengthen it: Make A World · Awestruck 3 Minutes · Spectrum Mapping
Network Design (34/100)
Exercises that strengthen it: Fun Trust-Building Group Game for Team Connection - Izzat... · Exciting Group Trust Activity for Building Confidence... · PFP: September 2005
Agenda Preview
The actual agenda we captured. Every block is classified by format and purpose. Open any block to see how we read it; the colored edge shows whether it is participant work, broadcast, logistics, or a showcase.
Room vs wrapper
4 percent of the 140 classified blocks put participants to work; the rest broadcast, show, or handle logistics. That mix is what drives the participation score.
Participant workBroadcastShowcaseLogistics
all eventBlog Track PostersPoster sessionShowcase+
Format · ShowcasePoster sessionPresenters display work; attendees browse and ask questions. Some interaction, not structured work.
Evidence basisOutcome inferred from formatInferred from format
all eventJournal Track PostersPoster sessionShowcase+
Format · ShowcasePoster sessionPresenters display work; attendees browse and ask questions. Some interaction, not structured work.
Evidence basisOutcome inferred from formatInferred from format
all eventWorkshopsWorkshopParticipant work+
Format · Participant workWorkshopParticipants work on a problem and produce something. The strongest signal of participation architecture.
Evidence basisParticipant work is implied by the formatInferred from format
2 p.mRegistration Desk 1RegistrationLogistics+
Format · LogisticsRegistrationCheck-in and logistics, not agenda content.
Evidence basisOutcome inferred from formatInferred from format
7:00 PMTHU 23 APRPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
all eventOpening RemarksOpeningOrientation+
Format · BroadcastOpeningFormat not classified from the source; treated as a broadcast block by default.
Evidence basisNo participant output visible from this rowRead from source, no work signal
9:00 AMInvited TalkPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:00 AMBreakBreakPacing+
Format · LogisticsBreakA pacing or recovery block between sessions.
Evidence basisOutcome inferred from formatInferred from format
10:30-11:40[10:30] Overthinking Reduction with Decoupled Rewards and Curriculum Data SchedulingPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:30[ ] Overthinking Reduction with Decoupled Rewards and Curriculum Data SchedulingPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:42[ ] Reducing Belief Deviation in Reinforcement Learning for Active Reasoning of LLM AgentsPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:54[ ] MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory AgentPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:06[ ] Verifying Chain-of-Thought Reasoning via Its Computational GraphPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:18[ ] Revela: Dense Retriever Learning via Language ModelingPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:30[ ] RAIN-Merging: A Gradient-Free Method to Enhance Instruction Following in Large Reasoning Models with Preserved Thinking FormatPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:30-11:52[10:30] Half-order Fine-Tuning for Diffusion Model: A Recursive Likelihood Ratio OptimizerPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:30[ ] Half-order Fine-Tuning for Diffusion Model: A Recursive Likelihood Ratio OptimizerPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:42[ ] Improving Diffusion Models for Class-imbalanced Training Data via Capacity ManipulationTrainingSkill building+
Format · Participant workTrainingGuided skill building where participants practice. Counts as participant work and learning transfer.
Evidence basisParticipant work is implied by the formatInferred from format
10:54[ ] Let Features Decide Their Own Solvers: Hybrid Feature Caching for Diffusion TransformersPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisSession topic references decisions or metrics; event-level follow-through is not shownRead from source
11:06[ ] DiffusionNFT: Online Diffusion Reinforcement with Forward ProcessPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:18[ ] Universal Inverse Distillation for Matching Models with Real-Data Supervision (No GANs)PresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:30[ ] GLASS Flows: Efficient Inference for Reward Alignment of Flow and Diffusion ModelsPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:42[ ] Neon: Negative Extrapolation From Self-Training Improves Image GenerationTrainingSkill building+
Format · Participant workTrainingGuided skill building where participants practice. Counts as participant work and learning transfer.
Evidence basisParticipant work is implied by the formatInferred from format
10:30-11:28[10:30] Mastering Sparse CUDA Generation through Pretrained Models and Deep Reinforcement LearningPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:30[ ] Mastering Sparse CUDA Generation through Pretrained Models and Deep Reinforcement LearningPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:42[ ] RefineStat: Efficient Exploration for Probabilistic Program SynthesisPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:54[ ] Huxley-G"odel Machine: Human-Level Coding Agent Development by an Approximation of the Optimal Self-Improving MachinePresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:06[ ] TileLang: Bridge Programmability and Performance in Modern Neural KernelsPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:18[ ] TabStruct: Measuring Structural Fidelity of Tabular DataPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:30-11:52[10:30] One for Two: A Unified Framework for Imbalanced Graph Classification via Dynamic Balanced PrototypePresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:30[ ] One for Two: A Unified Framework for Imbalanced Graph Classification via Dynamic Balanced PrototypePresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:42[ ] Compactness and Consistency: A Conjoint Framework for Deep Graph ClusteringPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:54[ ] Actions Speak Louder than Prompts: A Large-Scale Study of LLMs for Graph InferencePresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:06[ ] Multi-Domain Riemannian Graph Gluing for Building Graph Foundation ModelsPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:18[ ] Modality-free Graph In-context AlignmentPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:30[ ] Learning with Dual-level Noisy Correspondence for Multi-modal Entity AlignmentPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:42[ ] Exchangeability of GNN Representations with Applications to Graph RetrievalPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:30-11:52[10:30] Information Shapes Koopman RepresentationPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:30[ ] Information Shapes Koopman RepresentationPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:42[ ] On The Surprising Effectiveness of a Single Global Merging in Decentralized LearningPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:54[ ] Non-Convex Federated Optimization under Cost-Aware Client SelectionPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:06[ ] On the Wasserstein Geodesic Principal Component Analysis of probability measuresPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:18[ ] Fast Escape, Slow Convergence: Learning Dynamics of Phase Retrieval under Power-Law DataPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:30[ ] Hyperparameter Trajectory Inference with Conditional Lagrangian Optimal TransportPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:42[ ] Gaussian certified unlearning in high dimensions: A hypothesis testing approachPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:42[ ] Invisible Safety Threat: Malicious Finetuning for LLM via SteganographyPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:54[ ] The Shape of Adversarial Influence: Characterizing LLM Latent Spaces with Persistent HomologyPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:06[ ] Watch your steps: Dormant Adversarial Behaviors that Activate upon LLM FinetuningPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:18[ ] LLM Fingerprinting via Semantically Conditioned WatermarksPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:30[ ] Steering the Herd: A Framework for LLM-based Control of Social LearningPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
11:42[ ] Every Language Model Has a Forgery-Resistant SignaturePresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
10:30-1:00Causal Structure Learning in Hawkes Processes with Complex Latent Confounder NetworksPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
all eventMatching without Group Barrier for Heterogeneous Treatment Effect EstimationPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
all eventGDR-learners: Orthogonal Learning of Generative Models for Potential OutcomesPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
all eventQuery-Specific Causal Graph Pruning Under Tiered KnowledgePresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
all eventConditional Independent Component Analysis for Estimating Causal Structure with Latent VariablesPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
all eventMulti-Scale Diffusion-Guided Graph Learning with Power-Smoothing Random Walk Contrast for Multi-View ClusteringPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
all eventBridging ML and algorithms: comparison of hyperbolic embeddingsPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
all eventBoosting Open Set Recognition Performance through Modulated Representation LearningPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
all eventCross-Domain Lossy Compression via Rate- and Classification-Constrained Optimal TransportPresentationKnowledge transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisNo participant output visible from this rowRead from source, no work signal
all eventBlog Track PostersPoster SessionShowcase+
Format · ShowcasePoster SessionPresenters display work; attendees browse and ask questions. Some interaction, not structured work.
Evidence basisMediumRead from source
all eventJournal Track PostersPoster SessionShowcase+
Format · ShowcasePoster SessionPresenters display work; attendees browse and ask questions. Some interaction, not structured work.
Evidence basisMediumRead from source
all eventWorkshopsWorkshopCo Creation+
Format · Participant workWorkshopParticipants work on a problem and produce something. The strongest signal of participation architecture.
Evidence basisMediumRead from source
all eventBreak Expo Talk Panel Invited Talk Mentorship Oral Session Poster Session Reception Registration Desk Remarks Social Test Of Time Town Hall WorkshopPanelDeliberation+
Format · BroadcastPanelExperts discuss while the audience watches. Surfaces perspective but rarely creates participant work.
Evidence basisMediumRead from source
2 p.mRegistration Desk 1RegistrationLogistics+
Format · LogisticsRegistrationCheck-in and logistics, not agenda content.
Evidence basisMediumRead from source
7:00 PMTHU 23 APRUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
5:30 PMRemarks:UnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
all eventOpening RemarksOpening RemarksOrientation+
Format · BroadcastOpening RemarksFraming or welcome from the stage. Orients the room, not participatory.
Evidence basisMediumRead from source
9:00 AMInvited Talk:PresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:00 AMBreak:BreakWellbeing+
Format · LogisticsBreakA pacing or recovery block between sessions.
Evidence basisMediumRead from source
10:30 AM[Oral Session 1A LLMs and Reasoning [10:30-12:00] ](https://iclr.cc/virtual/2026/session/10011978)PresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:30-12:00[Oral Session 1A LLMs and Reasoning [ ] ](https://iclr.cc/virtual/2026/session/10011978)PresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:30-11:40[10:30] Overthinking Reduction with Decoupled Rewards and Curriculum Data SchedulingUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:30[ ] Overthinking Reduction with Decoupled Rewards and Curriculum Data SchedulingUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:42[ ] Reducing Belief Deviation in Reinforcement Learning for Active Reasoning of LLM AgentsUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:54[ ] MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory AgentUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:06[ ] Verifying Chain-of-Thought Reasoning via Its Computational GraphUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:18[ ] Revela: Dense Retriever Learning via Language ModelingUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:30[ ] RAIN-Merging: A Gradient-Free Method to Enhance Instruction Following in Large Reasoning Models with Preserved Thinking FormatUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
12:00 PM[Oral Session 1B Generative models I [10:30-12:00] ](https://iclr.cc/virtual/2026/session/10011979)PresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:30-12:00[Oral Session 1B Generative models I [ ] ](https://iclr.cc/virtual/2026/session/10011979)PresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:30-11:52[10:30] Half-order Fine-Tuning for Diffusion Model: A Recursive Likelihood Ratio OptimizerUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:30[ ] Half-order Fine-Tuning for Diffusion Model: A Recursive Likelihood Ratio OptimizerUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:42[ ] Improving Diffusion Models for Class-imbalanced Training Data via Capacity ManipulationTrainingSkill Building+
Format · Participant workTrainingGuided skill building where participants practice. Counts as participant work and learning transfer.
Evidence basisMediumRead from source
10:54[ ] Let Features Decide Their Own Solvers: Hybrid Feature Caching for Diffusion TransformersUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:06[ ] DiffusionNFT: Online Diffusion Reinforcement with Forward ProcessUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:18[ ] Universal Inverse Distillation for Matching Models with Real-Data Supervision (No GANs)UnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:30[ ] GLASS Flows: Efficient Inference for Reward Alignment of Flow and Diffusion ModelsUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:42[ ] Neon: Negative Extrapolation From Self-Training Improves Image GenerationTrainingSkill Building+
Format · Participant workTrainingGuided skill building where participants practice. Counts as participant work and learning transfer.
Evidence basisMediumRead from source
12:00 PM[Oral Session 1C Code generation and ML Systems [10:30-12:00] ](https://iclr.cc/virtual/2026/session/10011980)PresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:30-12:00[Oral Session 1C Code generation and ML Systems [ ] ](https://iclr.cc/virtual/2026/session/10011980)PresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:30-11:28[10:30] Mastering Sparse CUDA Generation through Pretrained Models and Deep Reinforcement LearningUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:30[ ] Mastering Sparse CUDA Generation through Pretrained Models and Deep Reinforcement LearningUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:42[ ] RefineStat: Efficient Exploration for Probabilistic Program SynthesisUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:54[ ] Huxley-G\"odel Machine: Human-Level Coding Agent Development by an Approximation of the Optimal Self-Improving MachineUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:06[ ] TileLang: Bridge Programmability and Performance in Modern Neural KernelsUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:18[ ] TabStruct: Measuring Structural Fidelity of Tabular DataUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
12:00 PM[Oral Session 1D Learning on graphs [10:30-12:00] ](https://iclr.cc/virtual/2026/session/10011981)PresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:30-12:00[Oral Session 1D Learning on graphs [ ] ](https://iclr.cc/virtual/2026/session/10011981)PresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:30-11:52[10:30] One for Two: A Unified Framework for Imbalanced Graph Classification via Dynamic Balanced PrototypeUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:30[ ] One for Two: A Unified Framework for Imbalanced Graph Classification via Dynamic Balanced PrototypeUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:42[ ] Compactness and Consistency: A Conjoint Framework for Deep Graph ClusteringUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:54[ ] Actions Speak Louder than Prompts: A Large-Scale Study of LLMs for Graph InferenceUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:06[ ] Multi-Domain Riemannian Graph Gluing for Building Graph Foundation ModelsUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:18[ ] Modality-free Graph In-context AlignmentUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:30[ ] Learning with Dual-level Noisy Correspondence for Multi-modal Entity AlignmentUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:42[ ] Exchangeability of GNN Representations with Applications to Graph RetrievalPresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
12:00 PM[Oral Session 1E Learning dynamics and optimization I [10:30-12:00] ](https://iclr.cc/virtual/2026/session/10011982)PresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:30-12:00[Oral Session 1E Learning dynamics and optimization I [ ] ](https://iclr.cc/virtual/2026/session/10011982)PresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:30-11:52[10:30] Information Shapes Koopman RepresentationPresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:30[ ] Information Shapes Koopman RepresentationPresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:42[ ] On The Surprising Effectiveness of a Single Global Merging in Decentralized LearningUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:54[ ] Non-Convex Federated Optimization under Cost-Aware Client SelectionUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:06[ ] On the Wasserstein Geodesic Principal Component Analysis of probability measuresUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:18[ ] Fast Escape, Slow Convergence: Learning Dynamics of Phase Retrieval under Power-Law DataUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:30[ ] Hyperparameter Trajectory Inference with Conditional Lagrangian Optimal TransportUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:42[ ] Gaussian certified unlearning in high dimensions: A hypothesis testing approachUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
12:00 PM[Oral Session 1F Privacy and security [10:30-12:00] ](https://iclr.cc/virtual/2026/session/10011983)PresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:30-12:00[Oral Session 1F Privacy and security [ ] ](https://iclr.cc/virtual/2026/session/10011983)PresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
10:30-11:52[10:30] Benchmarking Empirical Privacy Protection for Adaptations of Large Language ModelsUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:30[ ] Benchmarking Empirical Privacy Protection for Adaptations of Large Language ModelsUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:42[ ] Invisible Safety Threat: Malicious Finetuning for LLM via SteganographyUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
10:54[ ] The Shape of Adversarial Influence: Characterizing LLM Latent Spaces with Persistent HomologyUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:06[ ] Watch your steps: Dormant Adversarial Behaviors that Activate upon LLM FinetuningUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:18[ ] LLM Fingerprinting via Semantically Conditioned WatermarksUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:30[ ] Steering the Herd: A Framework for LLM-based Control of Social LearningUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
11:42[ ] Every Language Model Has a Forgery-Resistant SignatureUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
12:00 PM[Poster Session 1 Pavilion 3 [10:30-1:00] ](https://iclr.cc/virtual/2026/session/10021363)Poster SessionShowcase+
Format · ShowcasePoster SessionPresenters display work; attendees browse and ask questions. Some interaction, not structured work.
Evidence basisMediumRead from source
10:30-1:00[Poster Session 1 Pavilion 3 [ ] ](https://iclr.cc/virtual/2026/session/10021363)Poster SessionShowcase+
Format · ShowcasePoster SessionPresenters display work; attendees browse and ask questions. Some interaction, not structured work.
Evidence basisMediumRead from source
10:30-1:00Causal Structure Learning in Hawkes Processes with Complex Latent Confounder NetworksUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisMediumRead from source
all eventMatching without Group Barrier for Heterogeneous Treatment Effect EstimationUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisLowRead from source
all eventGDR-learners: Orthogonal Learning of Generative Models for Potential OutcomesUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisLowRead from source
all eventQuery-Specific Causal Graph Pruning Under Tiered KnowledgeUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisLowRead from source
all eventConditional Independent Component Analysis for Estimating Causal Structure with Latent VariablesUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisLowRead from source
all eventMulti-Scale Diffusion-Guided Graph Learning with Power-Smoothing Random Walk Contrast for Multi-View ClusteringUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisLowRead from source
all eventBridging ML and algorithms: comparison of hyperbolic embeddingsUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisLowRead from source
all eventBoosting Open Set Recognition Performance through Modulated Representation LearningPresentationKnowledge Transfer+
Format · BroadcastPresentationSpeakers present, the audience receives. Awareness only unless paired with practice or follow-up.
Evidence basisMediumRead from source
all eventAdaptive Width Neural NetworksUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisLowRead from source
all eventCross-Domain Lossy Compression via Rate- and Classification-Constrained Optimal TransportUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisLowRead from source
all eventRandom Controlled Differential EquationsUnknownUnknown+
Format · BroadcastUnknownFormat not classified from the source; treated as a broadcast block by default.
Evidence basisLowRead from source
The Full Reading
Why It Ranks This Way +
Calibrated from GES design 38/100 and verified 39/100, then capped for source appears to be a guide, directory, calendar, or ancillary page rather than a source agenda.
Reader Takeaway. For a reader, this is mainly a warning or source-evidence record: it reads as a lead or directory source, with the strongest visible signal in future-of-work fit and learning transfer and the biggest open question around follow through and participation architecture. The practical test is whether the published agenda connects the room to post-event continuation and evidence.
Strongest signals: Future-of-Work Fit, Learning Transfer, and Evidence Maturity. Weakest signals: Follow Through, Participation Architecture, and Network Design.
How This Agenda Could Improve +
- Add named owners, dates, implementation checkpoints, and a visible post-event continuation path.
- Turn passive airtime into participant work: practice, sensemaking, decisions, critique, or artifact creation.
- Replace generic networking blocks with designed introductions, ask-offer exchanges, peer groups, or bridge-building rituals.
Fastest next move: Add named owners, dated next steps, and a visible continuation path before treating the event as outcome-ready.
Role-Specific Reading +
Sponsor lens
Look beyond exposure. Strong sponsor value would show qualified interaction, problem work, buyer learning, customer evidence, or follow-up. The practical sponsor move is to look for structured introductions, buyer-seller fit, and relationship persistence.
Designer lens
The agenda is useful as a pattern sample from iclr.cc. Redesign attention should go first to the lowest-scoring pillars; in practice, turn the thinnest agenda blocks into participant work.
Executive lens
Treat the visible agenda as an operating plan. The executive move is to require owners, dates, and evidence before treating the event as strategic. If owners, proof, and follow-through are not visible, the public record does not yet prove strategic movement.
Aggregator lens
Treat the source URL as evidence, not decoration. The data-product move is to keep it as lead evidence unless a direct agenda source is also available before ranking or syndicating the record.
What GES Means Here +
The Gathering Effectiveness Score is a strict 0-100 public-evidence reading of the agenda across eight pillars. It rewards visible participant work, follow-through, transfer, network design, and proof mechanisms more than polish, speaker fame, attendance, or satisfaction.
Visible mechanisms: Participant work, Feedback, Network design, Learning transfer, Personalization.
Evidence boundary: Scores reflect visible agenda/source evidence and should not be read as proof of causal event impact.
Limitations, Score Caps, and Review Flags +
Limitations
- No visible follow-up, progress monitoring, or longitudinal tracking.
- Passive stage formats dominate the visible agenda.
- No baseline measurement is visible.
Score caps
- 28: Source appears to be a guide, directory, calendar, or ancillary page rather than a source agenda.
Review flags
- Fourth-loop score cap: Source appears to be a guide, directory, calendar, or ancillary page rather than a source agenda.
- URL path looks like an event calendar, directory, or list.
- Agenda is useful as a source record but weak as evidence of gathering effectiveness.
- No source-backed follow-up, validation, baseline, tracking, or impact evidence.
- No tracking, validation, feedback, or impact measurement found in the visible source text.
- Original category was unknown; publication category is inferred.
Is this proof the event worked? +
No. This is a strict public-evidence reading of the agenda. Proof would require baseline, comparison, follow-up, attribution, and impact evidence beyond the listing.
What should a reader inspect first? +
Start with the source URL, then compare the eight pillar scores against the agenda rows. The biggest opportunities usually sit in follow-through, evidence maturity, and participant work.
Why publish weak records? +
Weak records are part of the map. They show where public agendas still describe sessions and speakers more often than outcomes, commitments, transfer, or proof.
How should I use the rows? +
Read the agenda rows as the visible design trace: formats, purposes, and evidence labels show what the public source made inspectable, not everything that happened in the room. This is a source-grounded interpretation of the public agenda record, not a copy of the source, and not an endorsement of the event.
Where To Go Next
Compare this agenda against other Technology / AI / Startup events scored on the same eight pillars.