ICT framework applies JS divergence to token logits to select critical tokens for selective RLVR updates, claiming 4.58% average pass@4 gains on Qwen2.5 models across seven reasoning benchmarks.
Demystifying reason- ing dynamics with mutual information: Thinking tokens are information peaks in llm reasoning
9 Pith papers cite this work. Polarity classification is still indexing.
9
Pith papers citing it
representative citing papers
TESSY creates stylistically consistent synthetic data via teacher-student token interleaving, yielding 11.25% and 6.68% gains on code benchmarks where pure teacher data causes 3.25% and 10.02% drops.
The Stepwise Informativeness Assumption explains the correlation between LLM entropy dynamics and reasoning correctness by positing that correct traces accumulate answer-relevant information stepwise during generation.
Large reasoning models exhibit multilingual latent reasoning that is uneven across languages but internally consistent and English-centered.
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.
An attention-guided RL reward combined with diverse persuasion strategies produces higher attack success rates against large reasoning models than prior jailbreak methods.
CoRD uses collaborative multi-teacher step-wise decoding with perplexity-guided beam search to generate higher-quality Long-CoT data that lets smaller models reach near-teacher performance with less supervision.
COMPACT adaptively fuses multi-teacher CoT supervisions using graph-based consensus, mutual-information adaptability, and loss-based difficulty metrics to improve small language model reasoning performance while mitigating catastrophic forgetting.
ES-CoT shortens LLM chain-of-thought generation by tracking runs of identical step answers after linguistic markers, cutting tokens 16% on average while keeping accuracy comparable to full CoT across six datasets and three models.
citing papers explorer
-
Beyond Entropy: Learning from Token-Level Distributional Deviations for LLM Reasoning
ICT framework applies JS divergence to token logits to select critical tokens for selective RLVR updates, claiming 4.58% average pass@4 gains on Qwen2.5 models across seven reasoning benchmarks.
-
How to Fine-Tune a Reasoning Model? A Teacher-Student Cooperation Framework to Synthesize Student-Consistent SFT Data
TESSY creates stylistically consistent synthetic data via teacher-student token interleaving, yielding 11.25% and 6.68% gains on code benchmarks where pure teacher data causes 3.25% and 10.02% drops.
-
The Stepwise Informativeness Assumption: Why are Entropy Dynamics and Reasoning Correlated in LLMs?
The Stepwise Informativeness Assumption explains the correlation between LLM entropy dynamics and reasoning correctness by positing that correct traces accumulate answer-relevant information stepwise during generation.
-
Large Reasoning Models Are (Not Yet) Multilingual Latent Reasoners
Large reasoning models exhibit multilingual latent reasoning that is uneven across languages but internally consistent and English-centered.
-
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.
-
Attention-Guided Reward for Reinforcement Learning-based Jailbreak against Large Reasoning Models
An attention-guided RL reward combined with diverse persuasion strategies produces higher attack success rates against large reasoning models than prior jailbreak methods.
-
Distilling Long-CoT Reasoning through Collaborative Step-wise Multi-Teacher Decoding
CoRD uses collaborative multi-teacher step-wise decoding with perplexity-guided beam search to generate higher-quality Long-CoT data that lets smaller models reach near-teacher performance with less supervision.
-
"The Whole Is Greater Than the Sum of Its Parts": A Compatibility-Aware Multi-Teacher CoT Distillation Framework
COMPACT adaptively fuses multi-teacher CoT supervisions using graph-based consensus, mutual-information adaptability, and loss-based difficulty metrics to improve small language model reasoning performance while mitigating catastrophic forgetting.
-
Early Stopping Chain-of-thoughts in Large Language Models
ES-CoT shortens LLM chain-of-thought generation by tracking runs of identical step answers after linguistic markers, cutting tokens 16% on average while keeping accuracy comparable to full CoT across six datasets and three models.