Thinking Leakage: A Causal Audit of NoThink Post-Training in Hybrid Reasoning Models
cs.LG
Submitted: 2026-09-23
Updated: 2026-09-23
Terminology
Sources
- DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models
- Does On-Policy Distillation Really Distill? From Noisy Teacher to Self-Improvement
- OpenThoughts: Data Recipes for Reasoning Models
- Bootstrapping Exploration with Group-Level Natural Language Feedback in Reinforcement Learning
- Kimi K2.5: Visual Agentic Intelligence
- The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets
- MiniCPM4: Ultra-Efficient LLMs on End Devices
- Nemotron 3 Nano Omni: Efficient and Open Multimodal Intelligence
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- HybridFlow: A Flexible and Efficient RLHF Framework
- EasySteer: A Unified Framework for High-Performance and Extensible LLM Steering
- AGPO: Asymmetric Group Policy Optimization for Verifiable Reasoning and Search Ads Relevance at JD
- Qwen3 Technical Report
- Learning beyond Teacher: Generalized On-Policy Distillation with Reward Extrapolation
- Representation Engineering: A Top-Down Approach to AI Transparency
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks