Cover Image

Not All Negative Samples Are Equal: LLMs Learn Better from Plausible Reasoning

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Di, Zixiang, Han, Jinyi, Zhang, Shuo, Liao, Ying, Li, Zhi, Ji, Xiaofeng, Wang, Yongqi, Yang, Zheming, Gao, Ming, Li, Bingdong, Wang, Jie
Format:	Preprint
Published:	2026
Subjects:	Machine Learning Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Structured Reasoning for Large Language Models
by: Han, Jinyi, et al.
Published: (2026)

From Atoms to Chains: Divergence-Guided Reasoning Curriculum for Unlabeled LLM Domain Adaptation
by: Wang, Yongqi, et al.
Published: (2026)

CURE: Critical-Token-Guided Re-Concatenation for Entropy-Collapse Prevention
by: Li, Qingbin, et al.
Published: (2025)

Not All Tokens Are Created Equal: Query-Efficient Jailbreak Fuzzing for LLMs
by: Chen, Wenyu, et al.
Published: (2026)

Relation Reasoning with LLMs in Expensive Optimization
by: Lu, Ye, et al.
Published: (2026)

Not All Code Is Equal: A Data-Centric Study of Code Complexity and LLM Reasoning
by: Twist, Lukas, et al.
Published: (2026)

It's Morphing Time: Unleashing the Potential of Multiple LLMs via Multi-objective Optimization
by: Li, Bingdong, et al.
Published: (2024)

All Public Voices Are Equal, But Are Some More Equal Than Others to LLMs?
by: Kim, Sola, et al.
Published: (2026)

Does GCL Need a Large Number of Negative Samples? Enhancing Graph Contrastive Learning with Effective and Efficient Negative Sampling
by: Huang, Yongqi, et al.
Published: (2025)

Not All LLM Reasoners Are Created Equal
by: Hosseini, Arian, et al.
Published: (2024)

Not All Contexts Are Equal: Teaching LLMs Credibility-aware Generation
by: Pan, Ruotong, et al.
Published: (2024)

Not All Preferences Are Created Equal: Stability-Aware and Gradient-Efficient Alignment for Reasoning Models
by: Wu, Hui, et al.
Published: (2026)

Sample Smart, Not Hard: Correctness-First Decoding for Better Reasoning in LLMs
by: Li, Xueyan, et al.
Published: (2025)

ORCDF: An Oversmoothing-Resistant Cognitive Diagnosis Framework for Student Learning in Online Education Systems
by: Qian, Hong, et al.
Published: (2024)

All Cities are Equal: A Unified Human Mobility Generation Model Enabled by LLMs
by: Liu, Bo, et al.
Published: (2026)

Expensive Multi-Objective Bayesian Optimization Based on Diffusion Models
by: Li, Bingdong, et al.
Published: (2024)

Are All Pulsed Field Ablation Systems Equal?
by: Andrea Di Cori, et al.
Published: (2025)

Can LLMs Learn from Previous Mistakes? Investigating LLMs' Errors to Boost for Reasoning
by: Tong, Yongqi, et al.
Published: (2024)

Not All Parameters Are Created Equal: Smart Isolation Boosts Fine-Tuning Performance
by: Wang, Yao, et al.
Published: (2025)

Not All Correct Answers Are Equal: Why Your Distillation Source Matters
by: Tian, Xiaoyu, et al.
Published: (2025)

Pointer Networks Trained Better via Evolutionary Algorithms
by: Zhong, Muyao, et al.
Published: (2023)

TriSampler: A Better Negative Sampling Principle for Dense Retrieval
by: Yang, Zhen, et al.
Published: (2024)

Is Depth All You Need? An Exploration of Iterative Reasoning in LLMs
by: Wu, Zongqian, et al.
Published: (2025)

Adaptive Token-Weighted Differential Privacy for LLMs: Not All Tokens Require Equal Protection
by: Yu, Manjiang, et al.
Published: (2025)

Making Long-Context Language Models Better Multi-Hop Reasoners
by: Li, Yanyang, et al.
Published: (2024)

Plausible Reasoning and First-Order Plausible Logic
by: Billington, David
Published: (2026)

FedGA: Federated Learning with Gradient Alignment for Error Asymmetry Mitigation
by: Xiao, Chenguang, et al.
Published: (2024)

Not All Errors Are Equal: A Systematic Study of Error Propagation in Large Language Model Inference
by: Huang, Yafan, et al.
Published: (2026)

Under the Shadow of Babel: How Language Shapes Reasoning in LLMs
by: Wang, Chenxi, et al.
Published: (2025)

Conan-embedding: General Text Embedding with More and Better Negative Samples
by: Li, Shiyu, et al.
Published: (2024)

EduResearchBench: A Hierarchical Atomic Task Decomposition Benchmark for Full-Lifecycle Educational Research
by: Yue, Houping, et al.
Published: (2026)

Context-aware Diversity Enhancement for Neural Multi-Objective Combinatorial Optimization
by: Lu, Yongfan, et al.
Published: (2024)

Faithful or Just Plausible? Evaluating the Faithfulness of Closed-Source LLMs in Medical Reasoning
by: Afolabi, Halimat, et al.
Published: (2026)

Equal Access + Equal Opportunity = Success for All Learners
by: Agee, Schenell
Published: (2019)

All Beings Are Equal in Open Set Recognition
by: Li, Chaohua, et al.
Published: (2024)

Not All Pixels Are Equal: Confidence-Guided Attention for Feature Matching
by: Li, Dongyue
Published: (2025)

Toward Better EHR Reasoning in LLMs: Reinforcement Learning with Expert Attention Guidance
by: Fang, Yue, et al.
Published: (2025)

Not All Thoughts are Generated Equal: Efficient LLM Reasoning via Multi-Turn Reinforcement Learning
by: Ning, Yansong, et al.
Published: (2025)

SORM‐Enhanced Inverse Reliability Analysis for Geotechnical Multiobjective Reliability‐Based Design Optimization
by: Tao Wang, et al.
Published: (2024)

ResRL: Boosting LLM Reasoning via Negative Sample Projection Residual Reinforcement Learning
by: Lin, Zihan, et al.
Published: (2026)