22 papers
PrismaCoder: Pushing the Boundaries of Open Multimodal Code Intelligence through Decomposed Rubric Synthesis
Qiushi Sun, Jingyang Gong, Yang Liu +14
Multimodal code intelligence connects the symbolic world of programs with the visual world their execution creates. Interest in this capability is burgeoning, as models are increas…
Quality, Not Just Outcome: A Structural Scoring Framework for Selecting Coding-Agent Trajectories for Supervised Fine-Tuning
Sida Li, Yuanxin Liu, Shicheng Li +4
Open coding agents are commonly trained through supervised fine-tuning on interaction trajectories whose final patches pass the benchmark test suite. However, this outcome-level fi…
HARPO: Hallucination-Aware Reinforcement Learning for Faithful and Creative Language Generation
Tiezheng Yu, Yuxin Jiang, Jinpeng Li +4
Large Language Models (LLMs) are prone to generating hallucinated content, which compromises their reliability in knowledge-intensive tasks. To address this challenge without sacri…
EfficientAgent: What Makes KV Cache Offloading Work for Concurrent Agents?
Kunming Shao, Jierun Chen, Jiangnan Yu +7
LLM agents resend their whole conversation on every turn, and most of it was already processed on the previous turn. Serving systems avoid recomputing it by caching its key-value (…
OptiMAS: Automatically Optimize Multi-Agent System
Yuxin Cheng, Chang Liu, Hanxin Yu +9
Automated evolution of Multi-Agent Systems (MAS) holds significant potential for reducing the manual effort required to design and optimize LLM-based agent architectures. However,…
LEGO-RL: Harness-Native Reinforcement Learning for Coding Agents
Yiming Du, Yuxin Jiang, Tao Yuan +9
Reinforcement learning for coding agents increasingly relies on long-running agent harnesses to manage tool integration, repository contexts, and execution feedback. However, the n…
SWE-Review: Closing the Loop on Issue Resolution with Agentic Code Review
Ruoyu Wang, Jierun Chen, Shaowei Wang +7
Coding agents increasingly generate pull requests (PRs) for real-world software issues, yet one-shot PR generation remains open-loop: the PR is proposed without systematic review,…
What Makes Interaction Trajectories Effective for Training Terminal Agents?
Sidi Yang, Chaofan Tao, Jierun Chen +11
Stronger code agents are commonly assumed to be superior teachers for post-training, yet this assumption remains poorly disentangled from task difficulty, harness design, and stude…
QuantClaw: Precision Where It Matters for OpenClaw
Manyi Zhang, Ji-Fu Li, Zhongao Sun +5
Autonomous agent systems such as OpenClaw introduce significant efficiency challenges due to long-context inputs and multi-turn reasoning. This results in prohibitively high comput…
REAgent: Requirement-Driven LLM Agents for Software Issue Resolution
Shiqi Kuang, Zhao Tian, Kaiwei Lin +5
Issue resolution aims to automatically generate patches from given issue descriptions and has attracted significant attention with the rapid advancement of large language models (L…
FastCode: Fast and Cost-Efficient Code Understanding and Reasoning
Zhonghang Li, Zongwei Li, Yuxuan Chen +5
Repository-scale code reasoning is a cornerstone of modern AI-assisted software engineering, enabling Large Language Models (LLMs) to handle complex workflows from program comprehe…
Merging Beyond: Streaming LLM Updates via Activation-Guided Rotations
Yuxuan Yao, Haonan Sheng, Qingsong Lv +11
The escalating scale of Large Language Models (LLMs) necessitates efficient adaptation techniques. Model merging has gained prominence for its efficiency and controllability. Howev…
OVD: On-policy Verbal Distillation
Jing Xiong, Hui Shen, Shansan Gong +7
Knowledge distillation transfers reasoning capabilities from large teachers to efficient students. However, token-level on-policy distillation (OPD) constrains student exploration…
From Verifiable Dot to Reward Chain: Harnessing Verifiable Reference-based Rewards for Reinforcement Learning of Open-ended Generation
Yuxin Jiang, Yufei Wang, Qiyuan Zhang +6
Reinforcement learning with verifiable rewards (RLVR) succeeds in reasoning tasks (e.g., math and code) by checking the final verifiable answer (i.e., a verifiable dot signal). How…
SWE-Lego: Pushing the Limits of Supervised Fine-tuning for Software Issue Resolving
Chaofan Tao, Jierun Chen, Yuxin Jiang +11
We present SWE-Lego, a supervised fine-tuning (SFT) recipe designed to achieve state-ofthe-art performance in software engineering (SWE) issue resolving. In contrast to prevalent m…
Dream-VL & Dream-VLA: Open Vision-Language and Vision-Language-Action Models with Diffusion Language Model Backbone
Jiacheng Ye, Shansan Gong, Jiahui Gao +6
While autoregressive Large Vision-Language Models (VLMs) have achieved remarkable success, their sequential generation often limits their efficacy in complex visual planning and dy…
Memory-T1: Reinforcement Learning for Temporal Reasoning in Multi-session Agents
Yiming Du, Baojun Wang, Yifan Xiang +11
Temporal reasoning over long, multi-session dialogues is a critical capability for conversational agents. However, existing works and our pilot study have shown that as dialogue hi…
ADRA-Bank: A Modular Benchmark for Academic Deep Research Agents
Zhihan Guo, Feiyang Xu, Yifan Li +7
A surge in academic publications calls for automated deep research (DR) systems, but accurately evaluating them is still an open problem. First, existing benchmarks often focus nar…
HERMES: Towards Efficient and Verifiable Mathematical Reasoning in LLMs
Azim Ospanov, Zijin Feng, Jiacheng Sun +3
Informal mathematics has been central to modern large language model (LLM) reasoning, offering flexibility and efficient construction of arguments. However, purely informal reasoni…
ATTS: Asynchronous Test-Time Scaling via Conformal Prediction
Jing Xiong, Qiujiang Chen, Fanghua Ye +11
Large language models (LLMs) benefit from test-time scaling but are often hampered by high inference latency. Speculative decoding is a natural way to accelerate the scaling proces…
MMSearch-Plus: Benchmarking Provenance-Aware Search for Multimodal Browsing Agents
Xijia Tao, Yihua Teng, Xinxing Su +7
Existing multimodal browsing benchmarks often fail to require genuine multimodal reasoning, as many tasks can be solved with text-only heuristics without vision-in-the-loop verific…
S2D: Sorted Speculative Decoding For More Efficient Deployment of Nested Large Language Models
Parsa Kavehzadeh, Mohammadreza Pourreza, Mojtaba Valipour +5
Deployment of autoregressive large language models (LLMs) is costly, and as these models increase in size, the associated costs will become even more considerable. Consequently, di…