Image de couverture de livre

Not All Errors Are Created Equal: ASCoT Addresses Late-Stage Fragility in Efficient LLM Reasoning

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Zhang, Dongxu, Wu, Yujun, Sun, Yiding, Yang, Jinnan, Yang, Ning, Zhu, Jihua, Xin, Miao, Tian, Baoliang
Format:	Preprint
Publié:	2025
Sujets:	Computation and Language
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Not All LLM Reasoners Are Created Equal
par: Hosseini, Arian, et autres
Publié: (2024)

Chain-of-Thought Compression Should Not Be Blind: V-Skip for Efficient Multimodal Reasoning via Dual-Path Anchoring
par: Zhang, Dongxu, et autres
Publié: (2026)

Not All Queries Need Deep Thought: CoFiCot for Adaptive Coarse-to-fine Stateful Refinement
par: Zhang, Dongxu, et autres
Publié: (2026)

Not All Preferences Are Created Equal: Stability-Aware and Gradient-Efficient Alignment for Reasoning Models
par: Wu, Hui, et autres
Publié: (2026)

Not All Thoughts are Generated Equal: Efficient LLM Reasoning via Multi-Turn Reinforcement Learning
par: Ning, Yansong, et autres
Publié: (2025)

CMHANet: A Cross-Modal Hybrid Attention Network for Point Cloud Registration
par: Zhang, Dongxu, et autres
Publié: (2026)

Not All Preference Pairs Are Created Equal: A Recipe for Annotation-Efficient Iterative Preference Learning
par: Yang, Sen, et autres
Publié: (2024)

Not All Tokens Are Created Equal: Query-Efficient Jailbreak Fuzzing for LLMs
par: Chen, Wenyu, et autres
Publié: (2026)

Not All Options Are Created Equal: Textual Option Weighting for Token-Efficient LLM-Based Knowledge Tracing
par: Kim, JongWoo, et autres
Publié: (2024)

Selective Matching Losses -- Not All Scores Are Created Equal
par: Shamir, Gil I., et autres
Publié: (2025)

Not All Code Is Equal: A Data-Centric Study of Code Complexity and LLM Reasoning
par: Twist, Lukas, et autres
Publié: (2026)

Not All Entities are Created Equal: A Dynamic Anonymization Framework for Privacy-Preserving RAG
par: Zhu, Xinyuan, et autres
Publié: (2026)

Breaking the MoE LLM Trilemma: Dynamic Expert Clustering with Structured Compression
par: Zhu, Peijun, et autres
Publié: (2025)

Not All Explanations are Created Equal: Investigating the Pitfalls of Current XAI Evaluation
par: Shymanski, Joe, et autres
Publié: (2025)

Align then Adapt: Rethinking Parameter-Efficient Transfer Learning in 4D Perception
par: Sun, Yiding, et autres
Publié: (2026)

Not All Errors Are Equal: Investigation of Speech Recognition Errors in Alzheimer's Disease Detection
par: Kang, Jiawen, et autres
Publié: (2024)

Not All Errors Are Made Equal: A Regret Metric for Detecting System-level Trajectory Prediction Failures
par: Nakamura, Kensuke, et autres
Publié: (2024)

Device Context Protocol: A Compact, Safety-First Architecture for LLM-Driven Control of Constrained Devices
par: Yang, Dongxu
Publié: (2026)

Are All Marine Species Created Equal? Performance Disparities in Underwater Object Detection
par: Wille, Melanie, et autres
Publié: (2025)

Not All Parameters Are Created Equal: Smart Isolation Boosts Fine-Tuning Performance
par: Wang, Yao, et autres
Publié: (2025)

Not All Federated Learning Algorithms Are Created Equal: A Performance Evaluation Study
par: Baumgart, Gustav A., et autres
Publié: (2024)

Offline Policy Optimization with Posterior Sampling
par: Lin, Hongqiang, et autres
Publié: (2026)

PointCoT: A Multi-modal Benchmark for Explicit 3D Geometric Reasoning
par: Zhang, Dongxu, et autres
Publié: (2026)

Adversarial Samples Are Not Created Equal
par: Crawford, Jennifer, et autres
Publié: (2026)

AFBS:Buffer Gradient Selection in Semi-asynchronous Federated Learning
par: Lu, Chaoyi, et autres
Publié: (2025)

FedPSA: Modeling Behavioral Staleness in Asynchronous Federated Learning
par: Lu, Chaoyi, et autres
Publié: (2026)

All Entities are Not Created Equal: Examining the Long Tail for Ultra-Fine Entity Typing
par: Deshmukh, Advait, et autres
Publié: (2024)

Not All Deepfakes Are Created Equal: Triaging Audio Forgeries for Robust Deepfake Singer Identification
par: Salvi, Davide, et autres
Publié: (2025)

Not All RDF is Created Equal: Investigating RDF Load Times on Resource-Constrained Devices
par: Sowinski, Piotr, et autres
Publié: (2024)

Not All Bonds Are Created Equal: Dyadic Latent Class Models for Relational Event Data
par: Lakdawala, Rumana, et autres
Publié: (2025)

Not All Features Are Created Equal: A Mechanistic Study of Vision-Language-Action Models
par: Grant, Bryce, et autres
Publié: (2026)

Not All Layers Are Created Equal: Adaptive LoRA Ranks for Personalized Image Generation
par: Shenaj, Donald, et autres
Publié: (2026)

Not All Subhaloes Are Created Equal: Modelling the Diversity of Subhalo Density Profiles in TNG50
par: Heinze, Felix M., et autres
Publié: (2023)

Not All Errors Are Equal: A Systematic Study of Error Propagation in Large Language Model Inference
par: Huang, Yafan, et autres
Publié: (2026)

Not All Negative Samples Are Equal: LLMs Learn Better from Plausible Reasoning
par: Di, Zixiang, et autres
Publié: (2026)

Robust Answers, Fragile Logic: Probing the Decoupling Hypothesis in LLM Reasoning
par: Jiang, Enyi, et autres
Publié: (2025)

CompilerKV: Risk-Adaptive KV Compression via Offline Experience Compilation
par: Yang, Ning, et autres
Publié: (2026)

Not All Candidates are Created Equal: A Heterogeneity-Aware Approach to Pre-ranking in Recommender Systems
par: Tong, Pengfei, et autres
Publié: (2026)

Not All RAGs Are Created Equal: A Component-Wise Empirical Study for Software Engineering Tasks
par: Ke, Qiang, et autres
Publié: (2026)

Not All Frequencies Are Created Equal:Towards a Dynamic Fusion of Frequencies in Time-Series Forecasting
par: Zhang, Xingyu, et autres
Publié: (2024)