Papers by Aohan Zeng
Revisiting Parallel Context Windows: A Frustratingly Simple Alternative and Chain-of-Thought Deterioration (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing methods for extending the maximum context lengths of language models are lacking a strong baseline for in-context few-shot classification and on more challenging Chain-of-Thought reasoning, such as HotpotQA, deteriorate question miscomprehension and false inference. |
| Approach: | They propose to harness window-wise attention and positional embedding techniques to extend the maximum context lengths of language models. |
| Outcome: | The proposed method is able to extend the maximum context lengths of language models, e.g., 2048 for LLaMA, by harnessing window-wise attention and positional embedding techniques. |
Data Efficient RLVR via Off-Policy Influence Guidance (2026.acl-long)
Copied to clipboard
Erle Zhu, Dazhi Jiang, Yuan Wang, Xujun Li, Jiale Cheng, Yuxian Gu, Yilin Niu, Aohan Zeng, Jie Tang, Minlie Huang, Hongning Wang
| Challenge: | Existing data selection methods for RLVR are heuristic-based, lacking theoretical guarantees and generalizability. |
| Approach: | They propose an off-policy influence estimation method that approximates data influence using offline trajectories. |
| Outcome: | The proposed method reduces the computational cost of policy rollouts and improves storage and computation efficiency. |
MTP-RL: Acceleration of Reinforcement Learning Rollouts with Policy-Aligned Multi-Token Prediction (2026.findings-acl)
Copied to clipboard
| Challenge: | Reinforcement learning (RL) is widely applied to boost the performance of pretrained models, yet its training efficiency is severely constrained by rollout generation. |
| Approach: | They propose a framework that accelerates the rollout phase for diverse models by equipping a pipeline to equip the multi-layer parameter-sharing MTP for all models and an advantage-aware MTP optimization strategy. |
| Outcome: | The proposed framework achieves stable growth of acceptance length during RL training, and also accelerates RL rollouts, achieving an average 23.1%–55.3% reduction in rollout time compared to baselines. |
AgentTuning: Enabling Generalized Agent Abilities for LLMs (2024.findings-acl)
Copied to clipboard
| Challenge: | Open large language models (LLMs) with great performance in various tasks are far inferior to commercial models such as ChatGPT and GPT-4 when acting as agents to tackle complex tasks in the real world. |
| Approach: | They propose a method to enhance the agent capabilities of LLMs while maintaining their general abilities. |
| Outcome: | The AgentLM-70B is comparable to GPT-3.5-turbo on unseen agent tasks, demonstrating generalized agent capabilities. |
ChatGLM-Math: Improving Math Problem-Solving in Large Language Models with a Self-Critique Pipeline (2024.findings-emnlp)
Copied to clipboard
Yifan Xu, Xiao Liu, Xinghan Liu, Zhenyu Hou, Yueyan Li, Xiaohan Zhang, Zihan Wang, Aohan Zeng, Zhengxiao Du, Zhao Wenyi, Jie Tang, Yuxiao Dong
| Challenge: | Large language models (LLMs) have shown excellent mastering of human language but struggle in real-world applications that require mathematical problem-solving. |
| Approach: | They propose a pipeline to train a general Math-Critique model from the LLM itself to provide feedback signals and employ rejective fine-tuning and direct preference optimization over the Llm's own generations for data collection. |
| Outcome: | The proposed pipeline outperforms existing LLMs that could be two times larger. |
IF-CRITIC: Towards a Fine-Grained LLM Critic for Instruction-Following Evaluation (2026.acl-long)
Copied to clipboard
Bosi Wen, Yilin Niu, Cunxiang Wang, Pei Ke, Xiaoying Ling, Ying Zhang, Aohan Zeng, Hongning Wang, Minlie Huang
| Challenge: | Existing evaluation models for instruction-following have many shortcomings, such as substantial costs and unreliable assessments. |
| Approach: | They propose an LLM critic for fine-grained instruction-following evaluation using a checklist generator and a constraint-level preference optimization method. |
| Outcome: | The proposed model beats strong LLM-as-a-Judge baselines in evaluations under lower computational overhead compared to baselines. |
LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding (2024.acl-long)
Copied to clipboard
Yushi Bai, Xin Lv, Jiajie Zhang, Hongchang Lyu, Jiankai Tang, Zhidian Huang, Zhengxiao Du, Xiao Liu, Aohan Zeng, Lei Hou, Yuxiao Dong, Jie Tang, Juanzi Li
| Challenge: | Large language models (LLMs) can only handle texts a few thousand tokens long, limiting their applications on longer sequence inputs, such as books, reports, and codebases. |
| Approach: | They propose a bilingual, multi-task benchmark for long context understanding that extends context windows and more sophisticated memory mechanisms to improve models' long context capabilities. |
| Outcome: | The proposed model outperforms open-source models but struggles on longer contexts. |
CritiqueLLM: Towards an Informative Critique Generation Model for Evaluation of Large Language Model Generation (2024.acl-long)
Copied to clipboard
Pei Ke, Bosi Wen, Andrew Feng, Xiao Liu, Xuanyu Lei, Jiale Cheng, Shengyuan Wang, Aohan Zeng, Yuxiao Dong, Hongning Wang, Jie Tang, Minlie Huang
| Challenge: | Existing models for NLP evaluations lack the ability to generate informative critiques in pointwise grading and pairwise comparison especially without references. |
| Approach: | They propose a method which can acquire pointwise grading critiques with pseudo references and revise these critiques via multi-path prompting to obtain informative evaluation data in different tasks and settings. |
| Outcome: | The proposed method outperforms all open-source models and even GPT-4 in system-level correlations of pointwise grading. |