Static SFT and RL training for tool-use agents leads to performance drops under open-world distributional shifts across perception, interaction, reasoning and internalization; perturbation-augmented fine-tuning is proposed as mitigation.
hub
Llms can easily learn to reason from demonstrations structure, not content, is what matters!
18 Pith papers cite this work. Polarity classification is still indexing.
18
Pith papers citing it
hub tools
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
Adding explicit parent pointers to represent search tree structure in LLM reasoning traces (LinTree) improves task performance and search efficiency on Blocks World, grid Navigation, and Sokoban relative to implicit traces and LLM-heuristic search.
Cross-modal agreement between chain-of-thought and program-of-thought reasoning enables self-consistency with only two LLM samples, reducing sampling cost by 9.3x while improving accuracy.
The L2 norm of LLM hidden states signals reasoning intensity, with a theoretical bound on SAE feature activations, enabling three new test-time scaling techniques that boost performance.
Post-training quantization increases overthinking errors in reasoning models; a logit penalty on curated overthinking markers reduces CoT length 12-23% without accuracy loss.
Archer introduces response-level entropy normalization and differentiated clipping/KL regularization in RLVR to encourage exploration on reasoning tokens while stabilizing knowledge tokens, yielding gains in pass@1 and pass@K on reasoning benchmarks.
LongWriter-Zero applies RL from a base model with specialized rewards for length, quality, and structure to outperform SFT baselines and larger models on long-writing benchmarks.
LRMs exhibit complete accuracy collapse beyond certain puzzle complexities, with reasoning effort rising then declining, outperforming standard LLMs only on medium-complexity tasks.
High-entropy minority tokens drive RLVR gains, so restricting gradients to the top 20% maintains or improves performance over full updates on Qwen3 models, especially larger ones.
Reasoning in large output spaces proceeds via shortlisting then fine-grained reasoning; this characterization enables a mechanistic distillation strategy that outperforms standard distillation.
Absurd World automatically converts real-world problems into absurd yet logically coherent scenarios to test whether LLMs can reason without depending on familiar patterns.
Non-reasoning LLMs fail the equivalence class problem while reasoning LLMs perform better but remain incomplete, with difficulty peaking at phase transition for the former and maximum diameter for the latter.
A trajectory for student-LLM distillation is better when its tokens are surprising but still high-ranked, and the ratio of average rank to average surprisal (RSR) selects such trajectories better than existing metrics.
Reward-free online SFT on a model's own low-temperature samples improves math reasoning and roughly matches GRPO on six benchmarks.
A literature survey that introduces a taxonomy for LLM reasoning paradigms, analyzes methodological trends, and synthesizes failure modes from over 300 papers.
A new open SFT dataset for reasoning distillation lets coding models hit state-of-the-art scores on LiveCodeBench and CodeContests with supervised fine-tuning alone, outperforming RL-trained baselines.
ReasoningLens is a framework for hierarchical visualization, agentic auditing, and systemic profiling of long reasoning chains in large reasoning models.
This survey frames foundation agents using brain-inspired modular architectures and reviews challenges in evolution, collaboration, and safety.
citing papers explorer
-
Can Agents Generalize to the Open World? Unveiling the Fragility of Static Training in Tool Use
Static SFT and RL training for tool-use agents leads to performance drops under open-world distributional shifts across perception, interaction, reasoning and internalization; perturbation-augmented fine-tuning is proposed as mitigation.
-
LinTree: Improving LLM Reasoning with Explicitly Structured Search Histories
Adding explicit parent pointers to represent search tree structure in LLM reasoning traces (LinTree) improves task performance and search efficiency on Blocks World, grid Navigation, and Sokoban relative to implicit traces and LLM-heuristic search.
-
Self-Consistency from Only Two Samples: CoT-PoT Ensembling for Efficient LLM Reasoning
Cross-modal agreement between chain-of-thought and program-of-thought reasoning enables self-consistency with only two LLM samples, reducing sampling cost by 9.3x while improving accuracy.
-
The Tell-Tale Norm: $\ell_2$ Magnitude as a Signal for Reasoning Dynamics in Large Language Models
The L2 norm of LLM hidden states signals reasoning intensity, with a theoretical bound on SAE feature activations, enabling three new test-time scaling techniques that boost performance.
-
Quantized Reasoning Models Think They Need to Think Longer, but They Do Not
Post-training quantization increases overthinking errors in reasoning models; a logit penalty on curated overthinking markers reduces CoT length 12-23% without accuracy loss.
-
Stabilizing Knowledge, Promoting Reasoning: Dual-Token Constraints for RLVR
Archer introduces response-level entropy normalization and differentiated clipping/KL regularization in RLVR to encourage exploration on reasoning tokens while stabilizing knowledge tokens, yielding gains in pass@1 and pass@K on reasoning benchmarks.
-
LongWriter-Zero: Mastering Ultra-Long Text Generation via Reinforcement Learning
LongWriter-Zero applies RL from a base model with specialized rewards for length, quality, and structure to outperform SFT baselines and larger models on long-writing benchmarks.
-
The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity
LRMs exhibit complete accuracy collapse beyond certain puzzle complexities, with reasoning effort rising then declining, outperforming standard LLMs only on medium-complexity tasks.
-
Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning
High-entropy minority tokens drive RLVR gains, so restricting gradients to the top 20% maintains or improves performance over full updates on Qwen3 models, especially larger ones.
-
Characterize Then Distill: Mechanistic Reasoning in Large Output Spaces
Reasoning in large output spaces proceeds via shortlisting then fine-grained reasoning; this characterization enables a mechanistic distillation strategy that outperforms standard distillation.
-
Absurd World: A Simple Yet Powerful Method to Absurdify the Real-world for Probing LLM Reasoning Capabilities
Absurd World automatically converts real-world problems into absurd yet logically coherent scenarios to test whether LLMs can reason without depending on familiar patterns.
-
How Well Do LLMs Perform on the Simplest Long-Chain Reasoning Tasks: An Empirical Study on the Equivalence Class Problem
Non-reasoning LLMs fail the equivalence class problem while reasoning LLMs perform better but remain incomplete, with difficulty peaking at phase transition for the former and maximum diameter for the latter.
-
Which Reasoning Trajectories Teach Students to Reason Better? A Simple Metric of Informative Alignment
A trajectory for student-LLM distillation is better when its tokens are surprising but still high-ranked, and the ratio of average rank to average surprisal (RSR) selects such trajectories better than existing metrics.
-
A Model Can Help Itself: Reward-Free Self-Training for LLM Reasoning
Reward-free online SFT on a model's own low-temperature samples improves math reasoning and roughly matches GRPO on six benchmarks.
-
The Periodic Table of LLM Reasoning: A Structured Survey of Reasoning Paradigms, Methods, and Failure Modes
A literature survey that introduces a taxonomy for LLM reasoning paradigms, analyzes methodological trends, and synthesizes failure modes from over 300 papers.
-
OpenCodeReasoning: Advancing Data Distillation for Competitive Coding
A new open SFT dataset for reasoning distillation lets coding models hit state-of-the-art scores on LiveCodeBench and CodeContests with supervised fine-tuning alone, outperforming RL-trained baselines.
-
ReasoningLens: Hierarchical Visualization and Diagnostic Auditing for Large Reasoning Models
ReasoningLens is a framework for hierarchical visualization, agentic auditing, and systemic profiling of long reasoning chains in large reasoning models.
-
Advances and Challenges in Foundation Agents: From Brain-Inspired Intelligence to Evolutionary, Collaborative, and Safe Systems
This survey frames foundation agents using brain-inspired modular architectures and reviews challenges in evolution, collaboration, and safety.