Notes to Self: Can LLMs Benefit from Experiential Abstractions?

arXiv:2607.20372 · cs.CL · Submitted 2026-07-22 · Read on arXiv

cs.CL

Submitted: 2026-07-22

Updated: 2026-08-30

Code: https://github.com/ChangLiu-DrPatient/Notes-to-self

Terminology

• Abstraction — 142× in this paper · explained in 1 episode
• Trace — 33× in this paper · explained in 2 episodes
• GRPO — 25× in this paper · explained in 6 episodes
• Accuracy — 5× in this paper · explained in 1 episode
• Benchmark — 5× in this paper · explained in 2 episodes
• Reinforcement Learning — 5× in this paper · explained in 4 episodes
• Fine-Tuning — 3× in this paper · explained in 5 episodes
• Large Language Models (LLMs) — 3× in this paper · explained in 6 episodes
• Ablation — 2× in this paper · explained in 1 episode
• Reinforcement Learning (RL) — 2× in this paper · explained in 25 episodes
• Ablation study — 1× in this paper · explained in 1 episode
• Clustering — 1× in this paper · explained in 2 episodes
• Cosine Similarity — 1× in this paper · explained in 3 episodes
• Cross-Model Transfer — 1× in this paper · explained in 2 episodes
• Embeddings — 1× in this paper · explained in 2 episodes
• Iterative Refinement — 1× in this paper · explained in 3 episodes
• Test-time Scaling — 1× in this paper · explained in 2 episodes
• Transferability — 1× in this paper · explained in 4 episodes
• Verifier — 1× in this paper · explained in 2 episodes

Sources

• Measuring Mathematical Problem Solving With the MATH Dataset
• Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone
• Training Verifiers to Solve Math Word Problems
• RLAD: Training LLMs to Discover Abstractions for Solving Reasoning Problems
• Qwen2.5 Technical Report
• Metacognitive Reuse: Turning Recurring LLM Reasoning Into Concise Behaviors
• The Llama 3 Herd of Models
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• HybridFlow: A Flexible and Efficient RLHF Framework
• Gemma 3 Technical Report
• Gemma 2: Improving Open Language Models at a Practical Size
• SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning
• Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering