Copertina

EPiC: Towards Lossless Speedup for Reasoning Training through Edge-Preserving CoT Condensation

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Jia, Jinghan, Reisizadeh, Hadi, Fan, Chongyu, Baracaldo, Nathalie, Hong, Mingyi, Liu, Sijia
Natura:	Preprint
Pubblicazione:	2025
Soggetti:	Machine Learning
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Beyond SFT: Reinforcement Learning for Safer Large Reasoning Models with Better Reasoning Ability
di: Jia, Jinghan, et al.
Pubblicazione: (2025)

Reasoning Model Unlearning: Forgetting Traces, Not Just Answers, While Preserving Reasoning Skills
di: Wang, Changsheng, et al.
Pubblicazione: (2025)

EPiC: Efficient Video Camera Control Learning with Precise Anchor-Video Guidance
di: Wang, Zun, et al.
Pubblicazione: (2025)

Towards LLM Unlearning Resilient to Relearning Attacks: A Sharpness-Aware Minimization Perspective and Beyond
di: Fan, Chongyu, et al.
Pubblicazione: (2025)

Leak@$k$: Unlearning Does Not Make LLMs Forget Under Probabilistic Decoding
di: Reisizadeh, Hadi, et al.
Pubblicazione: (2025)

CyclicReflex: Improving Reasoning Models via Cyclical Reflection Token Scheduling
di: Fan, Chongyu, et al.
Pubblicazione: (2025)

BLUR: A Bi-Level Optimization Approach for LLM Unlearning
di: Reisizadeh, Hadi, et al.
Pubblicazione: (2025)

Defensive Unlearning with Adversarial Training for Robust Concept Erasure in Diffusion Models
di: Zhang, Yimeng, et al.
Pubblicazione: (2024)

WAGLE: Strategic Weight Attribution for Effective and Modular Unlearning in Large Language Models
di: Jia, Jinghan, et al.
Pubblicazione: (2024)

Rethinking Muon Beyond Pretraining: Spectral Failures and High-Pass Remedies for VLA and RLVR
di: Fan, Chongyu, et al.
Pubblicazione: (2026)

Downgrade to Upgrade: Optimizer Simplification Enhances Robustness in LLM Unlearning
di: Lang, Yicheng, et al.
Pubblicazione: (2025)

Parrot: A Training Pipeline Enhances Both Program CoT and Natural Language CoT for Reasoning
di: Jin, Senjie, et al.
Pubblicazione: (2025)

Simplicity Prevails: Rethinking Negative Preference Optimization for LLM Unlearning
di: Fan, Chongyu, et al.
Pubblicazione: (2024)

A Statistical Physics of Language Model Reasoning
di: Carson, Jack David, et al.
Pubblicazione: (2025)

Invariance Makes LLM Unlearning Resilient Even to Unanticipated Downstream Fine-Tuning
di: Wang, Changsheng, et al.
Pubblicazione: (2025)

CoT-based Synthesizer: Enhancing LLM Performance through Answer Synthesis
di: Zhang, Bohan, et al.
Pubblicazione: (2025)

C$^{2}$TC: A Training-Free Framework for Efficient Tabular Data Condensation
di: Xu, Sijia, et al.
Pubblicazione: (2026)

Reasoning as an Attack Surface: Adaptive Evolutionary CoT Jailbreaks for LLMs
di: Li, Jianan, et al.
Pubblicazione: (2026)

Challenging Forgets: Unveiling the Worst-Case Forget Sets in Machine Unlearning
di: Fan, Chongyu, et al.
Pubblicazione: (2024)

Faithfulness as Information Flow: Evaluating and Training Faithful Chain-of-Thought Reasoning
di: Jia, Jinghan, et al.
Pubblicazione: (2026)

From Explicit CoT to Implicit CoT: Learning to Internalize CoT Step by Step
di: Deng, Yuntian, et al.
Pubblicazione: (2024)

Distribution-Aligned Sequence Distillation for Superior Long-CoT Reasoning
di: Yan, Shaotian, et al.
Pubblicazione: (2026)

Mol-R1: Towards Explicit Long-CoT Reasoning in Molecule Discovery
di: Li, Jiatong, et al.
Pubblicazione: (2025)

VG-CoT: Towards Trustworthy Visual Reasoning via Grounded Chain-of-Thought
di: Lim, Byeonggeuk, et al.
Pubblicazione: (2026)

Improve Student's Reasoning Generalizability through Cascading Decomposed CoTs Distillation
di: Dai, Chengwei, et al.
Pubblicazione: (2024)

Exploring the Limitations of Mamba in COPY and CoT Reasoning
di: Ren, Ruifeng, et al.
Pubblicazione: (2024)

Evaluating LLM Reasoning Beyond Correctness and CoT
di: Abbasloo, Soheil
Pubblicazione: (2025)

Investigating CoT Monitorability in Large Reasoning Models
di: Yang, Shu, et al.
Pubblicazione: (2025)

Navigating Complexity: Toward Lossless Graph Condensation via Expanding Window Matching
di: Zhang, Yuchen, et al.
Pubblicazione: (2024)

Subspace Control: Turning Constrained Model Steering into Controllable Spectral Optimization
di: Huang, Yancheng, et al.
Pubblicazione: (2026)

Reconsidering Overthinking: Penalizing Internal and External Redundancy in CoT Reasoning
di: Hong, Jialiang, et al.
Pubblicazione: (2025)

D-SCoRE: Document-Centric Segmentation and CoT Reasoning with Structured Export for QA-CoT Data Generation
di: Zhou, Weibo, et al.
Pubblicazione: (2025)

Galaxy: A Cognition-Centered Framework for Proactive, Privacy-Preserving, and Self-Evolving LLM Agents
di: Bao, Chongyu, et al.
Pubblicazione: (2025)

One Token Embedding Is Enough to Deadlock Your Large Reasoning Model
di: Zhang, Mohan, et al.
Pubblicazione: (2025)

Distilling Long-CoT Reasoning through Collaborative Step-wise Multi-Teacher Decoding
di: Yun, Taewon, et al.
Pubblicazione: (2026)

The Synergy Dilemma of Long-CoT SFT and RL: Investigating Post-Training Techniques for Reasoning VLMs
di: Chen, Jierun, et al.
Pubblicazione: (2025)

Self-Verifying Reflection Helps Transformers with CoT Reasoning
di: Yu, Zhongwei, et al.
Pubblicazione: (2025)

CoT Vectors: Transferring and Probing the Reasoning Mechanisms of LLMs
di: Li, Li, et al.
Pubblicazione: (2025)

Efficient Long CoT Reasoning in Small Language Models
di: Wang, Zhaoyang, et al.
Pubblicazione: (2025)

Training on Documents About Monitoring Leads to CoT Obfuscation
di: Haskins, Reilly, et al.
Pubblicazione: (2026)