Good Papers

Showing papers from University of Chinese Academy of Sciences Show all papers

76%Highly rated
?Highly ratedVote to see the score
arXivInstitute of Computing TechnologU Chinese Academy of SciencesAlignment & preference optimization

On-Policy Parameter Update Direction Underlies Generalization in LLM Post-Training

On-policy methods continuously adjust parameter update directions, unlike consistent SFT updates; constraining SFT to these directions via OPSFT transfers on-policy generalization advantages to supervised fine-tuning.

Shufan Shen, Zhongni Hou, Junshu Sun, Yufei Zhang and 4 more

Published Sep 29, 2026 · 0 citations · ▲ 81 on Hugging Face

– ReadersNo votes yet
10/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 10 of 20 reviewers recommend it
lenient 4/5
medium 6/10
strict 0/5
57%Worth a look
?Worth a lookVote to see the score
Chinese Academy of SciencesInstitute of Information EngineeU Chinese Academy of SciencesBaiduEfficient attention & state-space models

Sparse Growing Transformer: Training-Time Sparse Depth Allocation via Progressive Attention Looping

Sparse Growing Transformer trains-time sparse depth allocation via progressive attention looping to improve efficiency.

Yao Chen, YiLong Chen, Yinqi Yang, Junyuan Shang and 8 more

Published 2026 · 0 citations

– ReadersNo votes yet
1/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 1 of 20 reviewers recommend it
lenient 1/5
medium 0/10
strict 0/5
54%Worth a look
?Worth a lookVote to see the score
Institute of Computing TechnologU Chinese Academy of SciencesTencentMulti-agent RL

Rhombus: Incentivizing Coordination in Parallel Thinking through Reinforcement Learning

Rhombus uses reinforcement learning to incentivize coordination in parallel thinking frameworks.

Ziyuan Nan, Qi Yi, Di Huang, Yutong Wu and 8 more

Published 2026 · 0 citations

100% Readers1 of 1 upvoted
0/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 0 of 20 reviewers recommend it
lenient 0/5
medium 0/10
strict 0/5
70%Highly rated
?Highly ratedVote to see the score
ACL 2025TsinghuaU Chinese Academy of SciencesZhipu AILong-context modeling

LongReward: Improving Long-context Large Language Models with AI Feedback

LongReward improves long-context LLMs by applying AI feedback to long-text instruction data via a multi-granularity reward model that evaluates both global coherence and local accuracy.

Jiajie Zhang, Zhongni Hou, Xin Lv, Shulin Cao and 6 more

Published 2025 · 4 citations

– ReadersNo votes yet
4/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 4 of 20 reviewers recommend it
lenient 2/5
medium 2/10
strict 0/5
45%Niche pick
?Niche pickVote to see the score
NeurIPS 2026SpotlightU Chinese Academy of SciencesU the Chinese Academy of ScienceInstitute of Computing Technolog3D vision & reconstruction

SceneScaffold: Active Scene-State Construction for Unified 3D Scene Understanding

Xiangqi Li, Libo Huang, Jiarui Zhao, Weilun Feng and 3 more

Sydney Poster Session 5, Thu, Dec 10, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

– ReadersNo votes yet
0/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 0 of 20 reviewers recommend it
lenient 0/5
medium 0/10
strict 0/5
45%Niche pick
?Niche pickVote to see the score
NeurIPS 2026U Chinese Academy of SciencesXiaohongshu IncU the Chinese Academy of ScienceSCIRInstitute of Software, Chinese AImitation learning

Your Teacher Can’t Help You Here: Combating Supervision Fidelity Decay in On-Policy Distillation

Yanjiang Liu, Jie Lou, Xinyan Guan, Yuqiu Ji and 6 more

Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

– ReadersNo votes yet
0/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 0 of 20 reviewers recommend it
lenient 0/5
medium 0/10
strict 0/5
45%Niche pick
?Niche pickVote to see the score
NeurIPS 2026Institute of Information EngineeKey Laboratory of Intelligent InChinese Academy of SciencesInstitute of Computing TechnologAlibaba GroupInstruction tuning

ToPA: Block-wise Toeplitz Adaptation for Expressive and Efficient Fine-Tuning

Sicong Li, Qianqian Xu, Zhiyong Yang, Zitai Wang and 3 more

Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

– ReadersNo votes yet
0/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 0 of 20 reviewers recommend it
lenient 0/5
medium 0/10
strict 0/5
57%Worth a look
?Worth a lookVote to see the score
NeurIPS 2026U the Chinese Academy of ScienceInstitute of Software, Chinese AU Chinese Academy of SciencesChinese Academy of Sciences, InsOffline RL

OGPO: Offline Goal-conditioned Policy Optimization for Recoverable Vision-Language-Action Models

Xule Gao, Xi Wang, Zehua Zang, Rui Wang and 2 more

Sydney Poster Session 5, Thu, Dec 10, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

– ReadersNo votes yet
1/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 1 of 20 reviewers recommend it
lenient 1/5
medium 0/10
strict 0/5
45%Niche pick
?Niche pickVote to see the score
NeurIPS 2026Institute of automation, ChineseU Chinese Academy of SciencesHuawei Technologies Ltd.Institute of Automation, ChineseUnified multimodal models

Attribution-Guided Shared-Private Decoupling for Noise-Reduced Audio-Visual Representation Learning

Linge Wang, Yingying Chen, Bingke Zhu, Lu Zhou and 3 more

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

– ReadersNo votes yet
0/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 0 of 20 reviewers recommend it
lenient 0/5
medium 0/10
strict 0/5
57%Worth a look
?Worth a lookVote to see the score
NeurIPS 2026SUN YAT-SEN UNIVERSITYAlibaba GroupKey Laboratory of Intelligent InU Chinese Academy of SciencesDiffusion models

Why DiT Models Underperform as Representation Learners without Long Skip Connections

Benyuan Meng, Yiliang Zhang, Jin-Wen Wu, Qianqian Xu and 2 more

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

– ReadersNo votes yet
1/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 1 of 20 reviewers recommend it
lenient 0/5
medium 0/10
strict 1/5
57%Worth a look
?Worth a lookVote to see the score
NeurIPS 2026Tianjin NormalU Chinese Academy of SciencesAutoML & architecture search

Treating Hyperparameters as Interventions: Task-Invariant Representation Learning for Transferable HPO

Mengyang Li, Ou Wu

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

– ReadersNo votes yet
1/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 1 of 20 reviewers recommend it
lenient 1/5
medium 0/10
strict 0/5
45%Niche pick
?Niche pickVote to see the score
NeurIPS 2026U the Chinese Academy of ScienceChinese Academy of SciencesVisual Information Processing anKey Laboratory of Intelligent InU Chinese Academy of SciencesOptimization

Pairwise AUC Optimization Needs Corrective Power: A Unified View

Jia Chen, Zhiyong Yang, Shilong Bao, Qianqian Xu and 1 more

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

– ReadersNo votes yet
0/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 0 of 20 reviewers recommend it
lenient 0/5
medium 0/10
strict 0/5
45%Niche pick
?Niche pickVote to see the score
NeurIPS 2026Institute of Computing TechnologU the Chinese Academy of ScienceU Chinese Academy of SciencesVIPL,ICT,Chinese academic of sciTest-time compute

Geometric Prompt-Trajectory Planning for Test-Time Scaling

Zhengqi Pei, Anran Zhang, Qingming Huang, Shuhui Wang

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

– ReadersNo votes yet
0/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 0 of 20 reviewers recommend it
lenient 0/5
medium 0/10
strict 0/5
45%Niche pick
?Niche pickVote to see the score
NeurIPS 2026Nanyang TechnologicalSun Yat-SenTencentTsinghua University, TsinghuaU Chinese Academy of SciencesDeep RL

DUDS: Dual-stage Data Selection for Efficient Reinforcement Learning with Verifiable Rewards

Hongling Zheng, Li Shen, Zichuan Lin, Jiafei Lyu and 5 more

Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

– ReadersNo votes yet
0/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 0 of 20 reviewers recommend it
lenient 0/5
medium 0/10
strict 0/5
45%Niche pick
?Niche pickVote to see the score
NeurIPS 2026U the Chinese Academy of ScienceTianjinComputer Network Information CenU Chinese Academy of SciencesInstruction tuning

DRIVE: Fine-tuning via Data Contribution- and Diversity-aware Weighting with Prior Regularization

qing liu, Xinrui Chen, Weiyao Zhu, Yi Du and 1 more

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

– ReadersNo votes yet
0/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 0 of 20 reviewers recommend it
lenient 0/5
medium 0/10
strict 0/5
88%Must read
?Must readVote to see the score
NeurIPS 2026Qilu University of TechnologyU the Chinese Academy of ScienceChina University of Petroleum-BeU Chinese Academy of SciencesWestlakeWeb & GUI agents

PAGER: Bridging the Semantic-Execution Gap in Point-Precise Geometric GUI Control

PAGER closes the semantic-execution gap for point-precise geometric GUI control via dependency-structured planning and pixel-level execution, achieving 4.1x higher task success than general baselines.

Jingxuan Wei, Xi Bai, Shan Liu, caijun jia and 7 more

Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026 · ▲ 15 on Hugging Face · Code ★ 3

– ReadersNo votes yet
15/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 15 of 20 reviewers recommend it
lenient 5/5
medium 7/10
strict 3/5
80%Must read
?Must readVote to see the score
NeurIPS 2026U the Chinese Academy of ScienceInstitute of Software, Chinese ABeijing University of AeronauticInstitute of Software, CASU Chinese Academy of SciencesRL for LLMs

Verifiable Environments Are LEGO Bricks: Recursive Composition for Reasoning Generalization

RACES recursively composes verifiable environments as LEGO bricks to scale RL reasoning training, boosting model performance on unseen benchmarks with far fewer base environments.

Hao Xiang, Qiaoyu Tang, Le Yu, Yaojie Lu and 7 more

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026 · ▲ 8 on Hugging Face

– ReadersNo votes yet
12/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 12 of 20 reviewers recommend it
lenient 4/5
medium 8/10
strict 0/5
72%Highly rated
?Highly ratedVote to see the score
NeurIPS 2026U the Chinese Academy of ScienceU Chinese Academy of SciencesKnowledge editing

MetaKE: Meta-Learning for Knowledge Editing Toward a Better Accuracy-Editability Trade-off

MetaKE unifies knowledge editing stages via bi-level optimization with structural gradient proxies, improving the accuracy-editability trade-off.

Shuxin Liu, Di Gao, Ou Wu

Sydney Poster Session 5, Thu, Dec 10, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

– ReadersNo votes yet
8/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 8 of 20 reviewers recommend it
lenient 3/5
medium 5/10
strict 0/5
78%Highly rated
?Highly ratedVote to see the score
NeurIPS 2026ZhejiangU Chinese Academy of SciencesTaobaoTaobao (China) Software Co., LtdAlibaba GroupLong-context modeling

StateTree: Enhancing Long-term Dialogue Reasoning via Reinforcement Learning

StateTree uses RL on tree-structured multi-session path-tracing tasks to train long-context dialogue reasoning, achieving up to +23.60% gains on LongMemEval and outperforming larger baselines while preserving short-context reasoning.

Naen Xu, Wanqing Cui, Yibo Hu, Shixin Hong and 4 more

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

– ReadersNo votes yet
11/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 11 of 20 reviewers recommend it
lenient 5/5
medium 5/10
strict 1/5
80%Must read
?Must readVote to see the score
NeurIPS 2026U Chinese Academy of SciencesJohns HopkinsPekingU the Chinese Academy of ScienceZhejiangVision-language models

dRAE: Representation Autoencoder with Hyper-Spherical Codes

Hyper-spherical quantization decouples semantics from magnitude via angular routing to prevent codebook collapse, enabling scalable discrete representation autoencoders with full codebook usage and high-fidelity reconstruction.

Tianren Ma, Lin Long, Chuyan Chen, Mu Zhang and 3 more

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026 · ▲ 10 on Hugging Face · Code ★ 10

– ReadersNo votes yet
12/20 AI panelreviewers recommend it

Readers and the AI panel: vote on this paper to see what they said.

Only vote on papers you've read. Sign in with GitHub to vote.

AI panel: 12 of 20 reviewers recommend it
lenient 5/5
medium 7/10
strict 0/5
Show 20 more papers