Reasoning as Compression: Unifying Budget Forcing via the Conditional Information Bottleneck
cs.LG
Submitted: 2026-03-09
Updated: 2026-11-04
Terminology
Sources
- L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learning
- OpenCodeReasoning: Advancing Data Distillation for Competitive Coding
- Compressed Chain of Thought: Efficient Reasoning Through Dense Representations
- Optimizing Length Compression in Large Reasoning Models
- IAPO: Information-Aware Policy Optimization for Token-Efficient Reasoning
- ThinkPrune: Pruning Long Chain-of-Thought of LLMs via Reinforcement Learning
- HAPO: Training Language Models to Reason Concisely via History-Aware Policy Optimization
- LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code
- Not All Turns Are Equally Hard: Adaptive Thinking Budgets For Efficient Multi-Turn Reasoning in Agents
- Overthinking Reduction with Decoupled Rewards and Curriculum Data Scheduling
- Revisiting LLM Reasoning via Information Bottleneck
- Steering LLM Thinking with Budget Guidance
- SelfBudgeter: Adaptive Token Allocation for Efficient LLM Reasoning
- Let's Verify Step by Step
- DLER: Doing Length pEnalty Right - Incentivizing More Intelligence per Token via Reinforcement Learning
- O1-Pruner: Length-Harmonizing Fine-Tuning for O1-Like Reasoning Pruning
- ODAR: Principled Adaptive Routing for LLM Reasoning via Active Inference
- Early Stopping Chain-of-thoughts in Large Language Models
- AIMO-2 Winning Solution: Building State-of-the-Art Mathematical Reasoning Models with OpenMathReasoning dataset
- GPQA: A Graduate-Level Google-Proof Q&A Benchmark
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis