15 papers
Beyond Distribution Matching: Selective Margin Preservation for Supervised Fine-Tuning
Yuxuan Wang, Guangxiang Zhao, Xinyu Tan +6
Supervised fine-tuning (SFT) changes logit margins among non-target tokens at teacher-forced prefixes. Matching a local reference distribution keeps those probabilities and also pe…
ScienceArena: Benchmarking LLMs on Latest Scientific Olympiad Competitions
Guangxiang Zhao, Qilong Shi, Xusen Xiao +13
Benchmark saturation and data contamination increasingly obscure genuine scientific reasoning in frontier LLMs. We introduce \textsc{ScienceArena}, an olympiad-style benchmark from…
ArborMem: Navigating Interaction States with Memory Forests
Zongwei Lv, Yuemeng Xu, Yilun Yao +8
Large language models increasingly serve as persistent conversational assistants, requiring memory that preserves relevant experience and maintains continuity across interactions.…
RealClawBench: Live OpenClaw Benchmarks from Real Developer-Agent Sessions
Zongwei Lv, Yaoming Li, Zhewen Tan +7
Agent benchmarks should reflect what users actually ask deployed agents to do, yet existing benchmarks often miss key realism properties of real developer-agent sessions. We introd…
A Primer in Post-Training Reasoning Data: What We Know About How It Works
Yaoming Li, Guangxiang Zhao, Qilong Shi +3
Post-training has become a primary driver of recent progress in large reasoning models, and reasoning data are often the key variable determining whether this stage succeeds. Work…
Harness-Bench: Evaluating Agent Harnesses across Models and Realistic Workflows
Yilun Yao, Xinyu Tan, Chao-Hsuan Liu +10
LLM agents are increasingly deployed as executable systems that use tools, modify workspaces, and produce concrete artifacts. In such workflows, performance depends not only on the…
Thinking with Reasoning Skills: Fewer Tokens, More Accuracy
Guangxiang Zhao, Qilong Shi, Xusen Xiao +3
Reasoning LLMs often spend substantial tokens on long intermediate reasoning traces (e.g., chain-of-thought) when solving new problems. We propose to summarize and store reusable r…
Beyond Parameter Arithmetic: Sparse Complementary Fusion for Distribution-Aware Model Merging
Weihong Lin, Lin Sun, Qilong Shi +6
Model merging has emerged as a promising paradigm for composing the capabilities of large language models by directly operating in weight space, enabling the integration of special…
Router Upcycling: Leveraging Mixture-of-Routers in Mixture-of-Experts Upcycling
Junfeng Ran, Guangxiang Zhao, Yuhan Wu +7
The Mixture-of-Experts (MoE) models have gained significant attention in deep learning due to their dynamic resource allocation and superior performance across diverse tasks. Howev…
Uncertainty Under the Curve: A Sequence-Level Entropy Area Metric for Reasoning LLM
Yongfu Zhu, Lin Sun, Guangxiang Zhao +2
In this work, we introduce Entropy Area Score (EAS), a simple yet effective metric to quantify uncertainty in the answer generation process of reasoning large language models (LLMs…
Evaluation is All You Need: Strategic Overclaiming of LLM Reasoning Capabilities Through Evaluation Design
Lin Sun, Weihong Lin, Jinzhu Wu +8
Reasoning models represented by the Deepseek-R1-Distill series have been widely adopted by the open-source community due to their strong performance in mathematics, science, progra…
TinyR1-32B-Preview: Boosting Accuracy with Branch-Merge Distillation
Lin Sun, Guangxiang Zhao, Xiaoqi Jian +18
The challenge of reducing the size of Large Language Models (LLMs) while maintaining their performance has gained significant attention. However, existing methods, such as model di…
Chain-of-Thought Matters: Improving Long-Context Language Models with Reasoning Path Supervision
Dawei Zhu, Xiyu Wei, Guangxiang Zhao +7
Recent advances in Large Language Models (LLMs) have highlighted the challenge of handling long-context tasks, where models need to reason over extensive input contexts to aggregat…
LongAttn: Selecting Long-context Training Data via Token-level Attention
Longyun Wu, Dawei Zhu, Guangxiang Zhao +5
With the development of large language models (LLMs), there has been an increasing need for significant advancements in handling long contexts. To enhance long-context capabilities…
Large Language Models Badly Generalize across Option Length, Problem Types, and Irrelevant Noun Replacements
Guangxiang Zhao, Saier Hu, Xiaoqi Jian +5
In this paper, we propose a ``Generalization Stress Test" to assess Large Language Models' (LLMs) generalization ability under slight and controlled perturbations, including option…