Elamur Research⌕
☰
home / papers / 2026 / Harnessing Consistency for Robust Test-Time LLM Ensemble

Harnessing Consistency for Robust Test-Time LLM Ensemble

Paper · 2026 · Zhichen Zeng, Qi Yu, Xiao Lin et al.

Summary

Test-time ensembles of large language models aim to combine the complementary strengths of different LLMs. However, they can be derailed by faulty signals. Mismatched tokenizers can cause tokens to be aligned incorrectly across vocabularies, and weak or unsure assistant models can pull fused predictions in the wrong direction. The authors find that such failures appear at two levels. At the token level, models strongly disagree on probabilities for specific tokens. At the model level, a model shows low confidence and diverges from its peers.

They propose CoRE, a plug-in reweighting scheme for token-level ensemble methods. It compares each assistant's projected probabilities against a uniform-average reference distribution. Tokens with large disparities are suppressed through a consistency function that acts like a low-pass filter. Models are weighted up when they combine low entropy with small divergence from the others. CoRE adds no extra inference cost. Across six benchmarks, several model combinations and multiple base ensemble methods, it reliably improves both accuracy and robustness.

Key points

  • Average gains: CoRE improves baseline ensemble methods by about 1.3% for Top-2 model ensembles and 2.8% for Top-3 ensembles, averaged over six benchmarks.
  • Motivating analysis: On 100 sampled NQ questions, misaligned tokens showed markedly larger probability disparity than correctly aligned ones (one-sided test, p ≈ 2.7 × 10⁻⁴⁴). Correct answers also showed lower entropy and higher summed token consistency than wrong ones.
  • Consistency functions on SAMSum: The authors tested three ways of turning disparity into a weight. The power-function variant gave the largest average gain (2.76), followed by RBF (2.35) and sigmoid (2.06). Results were fairly insensitive to the hyperparameters σ, α, β and γ.
  • Ablation: Restricting the token penalty to assistant models, instead of also applying it to the main model, yields better results; for instance, GAC gains +1.78 on PIQA, while EVA gains +1.66 on NQ. A case study ensembling OpenChat3.5 and InternLM2.5 shows CoRE fixing a misalignment in which "rine" was wrongly mapped to "_r".
  • Limitations and code: CoRE needs access to token-level logits, so it cannot be used with black-box APIs. Deciding when to ensemble and which models to combine remain open questions. Code is available at github.com/zhichenz98/CoRE-EACL26.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓University of Illinois Urbana-ChampaignUrbana · United States184 works · 69,372 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

🧭 Topics 1Research subjects and languages the work is about.Language models & pretraining · 3,320 works

Cite

Cite as Zhichen Zeng, Qi Yu, Xiao Lin, Ruizhong Qiu, Xuying Ning, Tianxin Wei, Yuchen Yan, Jingrui He and Hanghang Tong. 2026. Harnessing Consistency for Robust Test-Time LLM Ensemble. Findings of the Association for Computational Linguistics: EACL 2026.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.findings-eacl.182

@inproceedings{zeng2026harnessing,
  title = {{Harnessing Consistency for Robust Test-Time LLM Ensemble}},
  author = {Zhichen Zeng, Qi Yu, Xiao Lin, Ruizhong Qiu, Xuying Ning, Tianxin Wei, Yuchen Yan, Jingrui He and Hanghang Tong},
  booktitle = {Findings of the Association for Computational Linguistics: EACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-eacl.182},
  url = {https://doi.org/10.18653/v1/2026.findings-eacl.182}
}

References

85 references, 26 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Ahmed Abdulaal, Chen Jin, Nina Montaña-Brown, Aryo Pradipta Gema, Daniel C Castro, Daniel C Alexander, Philip Alexander Teare, Tom Diethe, Dino Oglic, and Amrutha Saseendran. Balancing act: Diversity and consistency in large language model ensembles. In The Thirteenth International Conference on Lea… find
  2. Xinyun Chen, Renat Aksitov, Uri Alon, Jie Ren, Kefan Xiao, Pengcheng Yin, Sushant Prakash, Charles Sutton, Xuezhi Wang, and Denny Zhou. 2023. Universal self-consistency for large language model generation. arXiv preprint arXiv:2311.17311.
  3. Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, and 1 others. 2021. Training verifiers to solve math word problems. arXiv preprint arXiv:2110.14168.
  4. Chengming Cui, Tianxin Wei, Ziyi Chen, Ruizhong Qiu, Zhichen Zeng, Zhining Liu, Xuying Ning, Duo Zhou, and Jingrui He. 2026. Adafuse: Adaptive ensemble decoding with test-time scaling for llms. arXiv preprint arXiv:2601.06022. — [arXiv:2601.06022](https://arxiv.org/abs/2601.06022) find
  5. Alexey Dosovitskiy. 2020. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929.
  6. Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, and 1 others. 2024. The llama 3 herd of models. arXiv e-prints, pages arXiv–2407.
  7. Yao Fu, Hao Peng, Litu Ou, Ashish Sabharwal, and Tushar Khot. 2023. Specializing smaller language models towards multi-step reasoning. In International Conference on Machine Learning, pages 10421–10430. PMLR.
  8. Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, and 1 others. 2023. Gpt-4 technical report. arXiv preprint arXiv:2303.08774.
  9. Bogdan Gliwa, Iwona Mochol, Maciej Biesek, and Aleksander Wawer. 2019. Samsum corpus: A human-annotated dialogue dataset for abstractive summarization. arXiv preprint arXiv:1911.12237.
  10. Pranjal Aggarwal, Aman Madaan, Yiming Yang, and 1 others. 2023. Let’s sample step by step: Adaptive-consistency for efficient reasoning and coding with llms. arXiv preprint arXiv:2305.11860. find
  11. Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, and 1 others. 2025. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv preprint arXiv:2501.12948.
  12. Yuntao Bai, Andy Jones, Kamal Ndousse, Amanda Askell, Anna Chen, Nova DasSarma, Dawn Drain, Stanislav Fort, Deep Ganguli, Tom Henighan, and 1 others. 2022. Training a helpful and harmless assistant with reinforcement learning from human feedback. arXiv preprint arXiv:2204.05862. find
  13. Yonatan Bisk, Rowan Zellers, Jianfeng Gao, Yejin Choi, and 1 others. 2020. Piqa: Reasoning about physical commonsense in natural language. In Proceedings Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. 2009. Measuring massive multitask language…
  14. Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. 2020. Measuring massive multitask language understanding. arXiv preprint arXiv:2009.03300.
  15. Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long short-term memory. Neural computation, 9(8):1735– 1780. find
  16. Yichong Huang, Xiaocheng Feng, Baohang Li, Yang Xiang, Hui Wang, Ting Liu, and Bing Qin. 2024. Ensemble learning for heterogeneous large language models with deep parallel collaboration. Advances in Neural Information Processing Systems, 37:119838– 119860. find
  17. Albert Q. Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lucile Saulnier, Lélio Renard Lavaud, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix,… find
  18. Dongfu Jiang, Xiang Ren, and Bill Yuchen Lin. 2023b. Llm-blender: Ensembling large language models with pairwise ranking and generative fusion. arXiv preprint arXiv:2306.02561.
  19. Mandar Joshi, Eunsol Choi, Daniel S Weld, and Luke Zettlemoyer. 2017. Triviaqa: A large scale distantly supervised challenge dataset for reading comprehension. arXiv preprint arXiv:1705.03551.
  20. Saurav Kadavath, Tom Conerly, Amanda Askell, Tom Henighan, Dawn Drain, Ethan Perez, Nicholas Schiefer, Zac Hatfield-Dodds, Nova DasSarma, Eli Tran-Johnson, and 1 others. 2022. Language models (mostly) know what they know. arXiv preprint arXiv:2207.05221. find
  21. Zhewei Kang, Xuandong Zhao, and Dawn Song. 2025. Scalable best-of-n selection for large language models via self-certainty. arXiv preprint arXiv:2502.18581. find
  22. Taku Kudo and John Richardson. 2018. Sentencepiece: A simple and language independent subword tokenizer and detokenizer for neural text processing. arXiv preprint arXiv:1808.06226.
  23. Tom Kwiatkowski, Jennimaria Palomaki, Olivia Redfield, Michael Collins, Ankur Parikh, Chris Alberti, Danielle Epstein, Illia Polosukhin, Jacob Devlin, Kenton Lee, and 1 others. 2019. Natural questions: a benchmark for question answering research. Transactions of the Association for Computational Lin…
  24. Yiwei Li, Peiwen Yuan, Shaoxiong Feng, Boyuan Pan, Xinglin Wang, Bin Sun, Heda Wang, and Kan Li. 2024a. Escape sky-high cost: Early-stopping self-consistency for multi-step reasoning. arXiv preprint arXiv:2401.10480. find
  25. Yunxuan Li, Yibing Du, Jiageng Zhang, Le Hou, Peter Grabowski, Yeqing Li, and Eugene Ie. 2024b. Improving multi-agent debate with sparse communication topology. arXiv preprint arXiv:2406.11776. find
  26. Zihao Li, Xiao Lin, Zhining Liu, Jiaru Zou, Ziwei Wu, Lecheng Zheng, Dongqi Fu, Yada Zhu, Hendrik Hamann, Hanghang Tong, and 1 others. 2025a. Language in the flow of time: Time-series-paired texts weaved into a unified temporal narrative. arXiv preprint arXiv:2502.08942. — [arXiv:2502.08942](https://arxiv.org/abs/2502.08942) find
  27. Zihao Li, Lecheng Zheng, Bowen Jin, Dongqi Fu, Baoyu Jing, Yikun Ban, Jingrui He, and Jiawei Han. 2025b. Can graph neural networks learn language with extremely weak text supervision? In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers),… find
  28. Mingfu Liang, Xi Liu, Rong Jin, Boyang Liu, Qiuling Suo, Qinghai Zhou, Song Zhou, Laming Chen, Hua Zheng, Zhiyuan Li, and 1 others. 2025. External large foundation model: How to efficiently serve trillions of parameters for online ads recommendation. In Companion Proceedings of the ACM on Web Confer… find
  29. Tian Liang, Zhiwei He, Wenxiang Jiao, Xing Wang, Yan Wang, Rui Wang, Yujiu Yang, Shuming Shi, and Zhaopeng Tu. 2023. Encouraging divergent thinking in large language models through multi-agent debate. arXiv preprint arXiv:2305.19118.
  30. Haokun Lin, Haobo Xu, Yichen Wu, Jingzhi Cui, Yingtao Zhang, Linzhan Mou, Linqi Song, Zhenan Sun, and Ying Wei. 2024a. Duquant: Distributing outliers via dual transformation makes stronger quantized llms. Advances in Neural Information Processing Systems, 37:87766–87800. find
  31. Haokun Lin, Haobo Xu, Yichen Wu, Ziyu Guo, Renrui Zhang, Zhichao Lu, Ying Wei, Qingfu Zhang, and Zhenan Sun. 2025a. Quantization meets dllms: A systematic study of post-training quantization for diffusion llms. arXiv preprint arXiv:2508.14896. — [arXiv:2508.14896](https://arxiv.org/abs/2508.14896) find
  32. Xiao Lin, Philip Li, Zhichen Zeng, Tingwei Li, Tianxin Wei, Xuying Ning, Gaotang Li, Yuzhong Chen, and Hanghang Tong. 2026. Alert: Zero-shot llm jailbreak detection via internal discrepancy amplification. arXiv preprint arXiv:2601.03600. — [arXiv:2601.03600](https://arxiv.org/abs/2601.03600) find
  33. Xiao Lin, Zhining Liu, Dongqi Fu, Ruizhong Qiu, and Hanghang Tong. 2024b. Backtime: Backdoor attacks on multivariate time series forecasting. Advances in Neural Information Processing Systems, 37:131344– 131368. find
  34. Xiao Lin, Zhichen Zeng, Tianxin Wei, Zhining Liu, Hanghang Tong, and 1 others. 2025b. Cats: Mitigating correlation shift for multivariate time series classification. arXiv preprint arXiv:2504.04283. — [arXiv:2504.04283](https://arxiv.org/abs/2504.04283) find
  35. Zhen Lin, Shubhendu Trivedi, and Jimeng Sun. 2023. Generating with confidence: Uncertainty quantification for black-box large language models. arXiv preprint arXiv:2305.19187. find
  36. Cong Liu, Xiaojun Quan, Yan Pan, Liang Lin, Weigang Wu, and Xu Chen. 2024a. Cool-fusion: Fuse large language models without training. arXiv preprint arXiv:2407.19807.
  37. Xiaolong Liu, Zhichen Zeng, Xiaoyi Liu, Siyang Yuan, Weinan Song, Mengyue Hang, Yiqun Liu, Chaofei Yang, Donghyun Kim, Wen-Yen Chen, and 1 others. 2024b. A collaborative ensemble framework for ctr prediction. arXiv preprint arXiv:2411.13700. — [arXiv:2411.13700](https://arxiv.org/abs/2411.13700) find
  38. Zhining Liu, Ze Yang, Xiao Lin, Ruizhong Qiu, Tianxin Wei, Yada Zhu, Hendrik Hamann, Jingrui He, and Hanghang Tong. 2025. Breaking silos: Adaptive model fusion unlocks better time series forecasting. arXiv preprint arXiv:2505.18442. — [arXiv:2505.18442](https://arxiv.org/abs/2505.18442) find
  39. Bo Lv, Chen Tang, Yanan Zhang, Xin Liu, Ping Luo, and Yue Yu. 2024a. Urg: A unified ranking and generation method for ensembling language models. In Findings of the Association for Computational Linguistics ACL 2024, pages 4421–4434. find
  40. Bo Lv, Chen Tang, Yanan Zhang, Xin Liu, Yue Yu, and Ping Luo. 2024b. Specfuse: Ensembling large language models via next-segment prediction. arXiv preprint arXiv:2412.07380. — [arXiv:2412.07380](https://arxiv.org/abs/2412.07380) find
  41. InternLM Team. 2023. Internlm: A multilingual language model with progressively enhanced capabilities. find
  42. Qwen Team. 2024. Qwen2 technical report. arXiv preprint arXiv:2407.10671, 2. find
  43. Selim Furkan Tekin, Fatih Ilhan, Tiansheng Huang, Sihao Hu, and Ling Liu. 2024. Llm-topla: Efficient llm ensemble by maximising diversity. arXiv preprint arXiv:2410.03953. find
  44. Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, and 1 others. 2023. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288.
  45. Fouad Trad and Ali Chehab. 2024. To ensemble or not: Assessing majority voting strategies for phishing detection with large language models. In International Conference on Intelligent Systems and Pattern Recognition, pages 158–173. Springer. find
  46. Costas Mavromatis, Petros Karypis, and George Karypis. 2024. Pack of llms: Model fusion at test-time via perplexity optimization. arXiv preprint arXiv:2404.11531. — [arXiv:2404.11531](https://arxiv.org/abs/2404.11531) find
  47. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. Advances in neural information processing systems, 30.
  48. Junichiro Niimi. 2025. A simple ensemble strategy for llm inference: Towards more stable text classification. In International Conference on Applications of Natural Language to Information Systems, pages 189–199. Springer. find
  49. Fanqi Wan, Xinting Huang, Deng Cai, Xiaojun Quan, Wei Bi, and Shuming Shi. 2024. Knowledge fusion of large language models. arXiv preprint arXiv:2401.10491. find
  50. Xuying Ning, Dongqi Fu, Tianxin Wei, Wujiang Xu, and Jingrui He. 2025. Graph4mm: Weaving multimodal learning with structural information. arXiv preprint arXiv:2510.16990. — [arXiv:2510.16990](https://arxiv.org/abs/2510.16990) find
  51. Guan Wang, Sijie Cheng, Xianyuan Zhan, Xiangang Li, Sen Song, and Yang Liu. 2023. Openchat: Advancing open-source language models with mixed-quality data. arXiv preprint arXiv:2309.11235.
  52. Rico Sennrich, Barry Haddow, and Alexandra Birch. 2016. Neural machine translation of rare words with subword units. In Proceedings of the 54th annual meeting of the association for computational linguistics (volume 1: long papers), pages 1715–1725.
  53. Junlin Wang, Jue Wang, Ben Athiwaratkun, Ce Zhang, and James Zou. 2024. Mixture-of-agents enhances large language model capabilities. arXiv preprint arXiv:2406.04692. find
  54. Chenglei Si, Weijia Shi, Chen Zhao, Luke Zettlemoyer, and Jordan Boyd-Graber. 2023. Getting more out of mixture of language model reasoning experts. arXiv preprint arXiv:2305.14628. find
  55. Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc Le, Ed Chi, Sharan Narang, Aakanksha Chowdhery, and Denny Zhou. 2022. Self-consistency improves chain of thought reasoning in language models. arXiv preprint arXiv:2203.11171.
  56. Amir Taubenfeld, Tom Sheffer, Eran Ofek, Amir Feder, Ariel Goldstein, Zorik Gekhman, and Gal Yona. 2025. Confidence improves self-consistency in llms. arXiv preprint arXiv:2502.06233. find
  57. Alexander Wei, Nika Haghtalab, and Jacob Steinhardt. 2023. Jailbroken: How does llm safety training fail? Advances in Neural Information Processing Systems, 36:80079–80110.
  58. Gemini Team, Rohan Anil, Sebastian Borgeaud, Jean-Baptiste Alayrac, Jiahui Yu, Radu Soricut, Johan Schalkwyk, Andrew M Dai, Anja Hauth, Katie Millican, and 1 others. 2023. Gemini: a family of highly capable multimodal models. arXiv preprint arXiv:2312.11805.
  59. Yonghui Wu, Mike Schuster, Zhifeng Chen, Quoc V Le, Mohammad Norouzi, Wolfgang Macherey, Maxim Krikun, Yuan Cao, Qin Gao, Klaus Macherey, and 1 others. 2016. Google’s neural machine translation system: Bridging the gap between human and machine translation. arXiv preprint arXiv:1609.08144. find
  60. Haobo Xu, Yuchen Yan, Dingsu Wang, Zhe Xu, Zhichen Zeng, Tarek F Abdelzaher, Jiawei Han, and Hanghang Tong. 2024a. Slog: An inductive spectral graph neural network beyond polynomial filter. In Forty-first International Conference on Machine Learning. find
  61. Qi Yu, Zhichen Zeng, Yuchen Yan, Zhining Liu, Baoyu Jing, Ruizhong Qiu, Ariful Azad, and Hanghang Tong. 2025a. Planetalign: A comprehensive python library for benchmarking network alignment. arXiv preprint arXiv:2505.21366. — [arXiv:2505.21366](https://arxiv.org/abs/2505.21366) find
  62. Yangyifan Xu, Jianghao Chen, Junhong Wu, and Jiajun Zhang. 2024b. Hit the sweet spot! span-level ensemble for large language models. arXiv preprint arXiv:2409.18583.
  63. Qi Yu, Zhichen Zeng, Yuchen Yan, Lei Ying, R Srikant, and Hanghang Tong. 2025b. Joint optimal transport and embedding for network alignment. In Proceedings of the ACM on Web Conference 2025, pages 2064–2075. find
  64. Yangyifan Xu, Jinliang Lu, and Jiajun Zhang. 2024c. Bridging the gap between different vocabularies for llm ensemble. arXiv preprint arXiv:2404.09492. find
  65. Yuchen Yan, Yuzhong Chen, Huiyuan Chen, Xiaoting Li, Zhe Xu, Zhichen Zeng, Lihui Liu, Zhining Liu, and Hanghang Tong. 2024a. Thegcn: Temporal heterophilic graph convolutional network. arXiv preprint arXiv:2412.16435. — [arXiv:2412.16435](https://arxiv.org/abs/2412.16435) find
  66. Yuchen Yan, Yongyi Hu, Qinghai Zhou, Lihui Liu, Zhichen Zeng, Yuzhong Chen, Menghai Pan, Huiyuan Chen, Mahashweta Das, and Hanghang Tong. 2024b. Pacer: Network embedding from positional to structural. In Proceedings of the ACM Web Conference 2024, pages 2485–2496. find
  67. Yuchen Yan, Aakash Kolekar, Sahika Genc, Wenju Xu, Edward W Huang, Anirudh Srinivasan, Mukesh Jain, Qi He, and Hanghang Tong. 2025. To answer or not to answer (taona): A robust textual graph understanding and question answering approach. In Findings of the Association for Computational Linguistics:… find
  68. Yuchen Yan, Lihui Liu, Yikun Ban, Baoyu Jing, and Hanghang Tong. 2021a. Dynamic knowledge graph alignment. In Proceedings of the AAAI conference on artificial intelligence, volume 35, pages 4564–4572. find
  69. Yuchen Yan, Si Zhang, and Hanghang Tong. 2021b. Bright: A bridging algorithm for network alignment. In Proceedings of the web conference 2021, pages 3907–3917. find
  70. Yuchen Yan, Qinghai Zhou, Jinning Li, Tarek Abdelzaher, and Hanghang Tong. 2022. Dissecting cross-layer dependency inference on multi-layered interdependent networks. In Proceedings of the 31st ACM International Conference on Information & Knowledge Management, pages 2341–2351. find
  71. Yuxuan Yao, Han Wu, Mingyang Liu, Sichun Luo, Xiongwei Han, Jie Liu, Zhijiang Guo, and Linqi Song. 2024. Determine-then-ensemble: Necessity of top-k union for large language model ensembling. arXiv preprint arXiv:2410.03777. — [arXiv:2410.03777](https://arxiv.org/abs/2410.03777) find
  72. Hyunsik Yoo, Zhichen Zeng, Jian Kang, Ruizhong Qiu, David Zhou, Zhining Liu, Fei Wang, Charlie Xu, Eunice Chan, and Hanghang Tong. 2024. Ensuring user-side fairness in dynamic recommender systems. In Proceedings of the ACM Web Conference 2024, pages 3667–3678. find
  73. Yao-Ching Yu, Chun-Chih Kuo, Ziqi Ye, Yu-Cheng Chang, and Yueh-Se Li. 2024. Breaking the ceiling of the llm community by treating token generation as a classification for ensembling. arXiv preprint arXiv:2406.12585. find
  74. Zhichen Zeng, Wenxuan Bao, Xiao Lin, Ruizhong Qiu, Tianxin Wei, Xuying Ning, Yuchen Yan, Chen Luo, Monica Xiao Cheng, Jingrui He, and 1 others. 2026. Subspace alignment for vision-language model test-time adaptation. arXiv preprint arXiv:2601.08139. — [arXiv:2601.08139](https://arxiv.org/abs/2601.08139) find
  75. Zhichen Zeng, Boxin Du, Si Zhang, Yinglong Xia, Zhining Liu, and Hanghang Tong. 2024a. Hierarchical multi-marginal optimal transport for network alignment. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 16660–16668. find
  76. Zhichen Zeng, Mengyue Hang, Xiaolong Liu, Xiaoyi Liu, Xiao Lin, Ruizhong Qiu, Tianxin Wei, Zhining Liu, Siyang Yuan, Chaofei Yang, and 1 others. 2025a. Hierarchical lora moe for efficient ctr model scaling. arXiv preprint arXiv:2510.10432. — [arXiv:2510.10432](https://arxiv.org/abs/2510.10432) find
  77. Zhichen Zeng, Xiaolong Liu, Mengyue Hang, Xiaoyi Liu, Qinghai Zhou, Chaofei Yang, Yiqun Liu, Yichen Ruan, Laming Chen, Yuxin Chen, and 1 others. 2025b. Interformer: Effective heterogeneous interaction learning for click-through rate prediction. In Proceedings of the 34th ACM International Conference… find
  78. Zhichen Zeng, Ruizhong Qiu, Wenxuan Bao, Tianxin Wei, Xiao Lin, Yuchen Yan, Tarek F Abdelzaher, Jiawei Han, and Hanghang Tong. 2025c. Pave your own path: Graph gradual domain adaptation on fused gromov-wasserstein geodesics. arXiv preprint arXiv:2505.12709. — [arXiv:2505.12709](https://arxiv.org/abs/2505.12709) find
  79. Zhichen Zeng, Ruizhong Qiu, Zhe Xu, Zhining Liu, Yuchen Yan, Tianxin Wei, Lei Ying, Jingrui He, and Hanghang Tong. 2024b. Graph mixup on approximate gromov–wasserstein geodesics. In Forty-first International Conference on Machine Learning. find
  80. Zhichen Zeng, Si Zhang, Yinglong Xia, and Hanghang Tong. 2023a. Parrot: Position-aware regularized optimal transport for network alignment. In Proceedings of the ACM web conference 2023, pages 372–382. find
  81. Zhichen Zeng, Ruike Zhu, Yinglong Xia, Hanqing Zeng, and Hanghang Tong. 2023b. Generative graph dictionary learning. In International Conference on Machine Learning, pages 40749–40769. PMLR. find
  82. Yuheng Zhang, Dian Yu, Tao Ge, Linfeng Song, Zhichen Zeng, Haitao Mi, Nan Jiang, and Dong Yu. 2025. Improving llm general preference alignment via optimistic online mirror descent. arXiv preprint arXiv:2502.16852. — [arXiv:2502.16852](https://arxiv.org/abs/2502.16852) find
  83. Lecheng Zheng, Baoyu Jing, Zihao Li, Hanghang Tong, and Jingrui He. 2024. Heterogeneous contrastive learning for foundation models and beyond. In Proceedings of the 30th ACM SIGKDD conference on knowledge discovery and data mining, pages 6666– 6676. find
  84. Kaijie Zhu, Jindong Wang, Jiaheng Zhou, Zichen Wang, Hao Chen, Yidong Wang, Linyi Yang, Wei Ye, Yue Zhang, Neil Gong, and 1 others. 2023. Promptrobust: Towards evaluating the robustness of large language models on adversarial prompts. In Proceedings of the 1st ACM workshop on large AI systems and mo…
  85. Andy Zou, Zifan Wang, Nicholas Carlini, Milad Nasr, J Zico Kolter, and Matt Fredrikson. 2023. Universal and transferable adversarial attacks on aligned language models. arXiv preprint arXiv:2307.15043.

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/zeng-2026-harnessing-consistency-robust-test-time/