acceptodds: Which papers will get accepted at ICLR 2027?
Abstract
A prediction market on peer review. Each paper has a market on its decision (accept or reject), priced by researchers who stake $rep, an in app currency, on what they expect. Everyone who signs up receives 1,000 $rep to trade with. Prices are probabilities, and every market settles when the venue publishes its decisions. It is a game made for fun, and has no connection to ICLR, OpenReview or any other organisation. acceptodds is an open source project, open to contributions, at github.com/benedict-armstrong/acceptodds.1
[PDF]Revisiting Log-Transformation in Deep Regression: Rectifying Metric-Warped Interpolation Bias via Transient Variance Guidance
[PDF]Why This Expert? A Feature-Level Analysis of Routing in Mixture-of-Experts with Sparse Autoencoders
[PDF]Bridging the Completeness Gap in GraphRAG: A Unified GNN Enhancement Layer for Retrieval and Generation
[PDF]See What I Need, Dream What I Want: A Navigation Foundation Model on a Self-Constructed Representation Space
[PDF]Synchronizing Work with Results in Long-Horizon Computer-Use Agents via Demand-Linked Execution
[PDF]Do Multi-Turn Conversational Capabilities Carry Over from Text to Speech? A Parallel Multilingual Benchmark
[PDF]When Recovery Creates Resistance: Instrumental Shutdown Avoidance from Gridworlds to Language-Model Agents
[PDF]Compress What You See, Not What You Say: Anchored Context Distillation for Latent-Observation Software Engineering Agents
[PDF]Sin-SMFL: An Ablation Study of Modulation, Class Balancing, and Gain Calibration in Detection Losses
[PDF]Bits Under ZK-LLM: Evaluating Zero-Knowledge-Friendly Quantization for Verifiable Private LLM Inference
1 If this project is interesting to you, reach out at [email protected].