JJ
Jiantao Jiao
cs.LGcs.AIstat.MLcs.ITmath.ITmath.STstat.THcs.CLcs.CReess.SP
On Valency
published · living versionsW_472urbcs·v1 · currentpublished
FrontierCS 2: A Benchmark for Computer Science R&D
with Qiuyang Mang, Wenhao Chai, Runyuan He, Dongruixuan Li +56
1 version
Preprints & journals
100 papers in the corpus · 2011–2026FrontierCS 2: A Benchmark for Computer Science R&D2610.29932v1 · Qiuyang Mang, Wenhao Chai, Runyuan He et al.2026 · 0 citationsarXivon Valency
A Good Self-Teacher Meets the Student Where They Are: Joint On-Policy Learning and Teaching2610.10447v1 · Randy Ardywibowo, Arnav Dalal, Jiantao Jiao2026 · 0 citationsarXiv
Before They Can Solve: Predicting Post-Training Coding-Agent Performance from Base Models2610.10478v1 · Tan Yu, Alexander Bukharin, Khushi Bhardwaj et al.2026 · 0 citationsarXiv
Where the Model Changes Its Mind: Hindsight-Divergence Localization for Efficient Reinforcement Learning with Verifiable Rewards2609.36864v1 · Fanchao Chen, Hengyu Fu, Shivaram Venkataraman et al.2026 · 0 citationsarXiv
SWE-Serve: Benchmarking Agentic Engineering For Production Inference Serving2609.26777v1 · Jennifer Williams, Dave Farris, Jeff Farris et al.2026 · 0 citationsarXiv
Coresets Before Score Sets: Evaluation-Unsupervised Prompt Subset Selection for LLM Benchmarks2607.09739v1 · Jihan Yao, Gantavya Bhatt, Arnav Das et al.2026 · 0 citationsarXiv
DiscoLoop: Looping Discrete Embeddings and Continuous Hidden States for Multi-hop Reasoning2607.00341v2 · Hengyu Fu, Tianyu Guo, Zixuan Wang et al.2026 · 0 citationsarXiv
Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning2606.15007v1 · NVIDIA: Aaron Blakeman, Aaron Thomas, Aastha Jhunjhunwala et al.2026 · 0 citationsarXiv
DiPOD: Diffusion Policy Optimization without Drifting Apart2606.13795v2 · Haozhe Jiang, Haiwen Feng, Pieter Abbeel et al.2026 · 0 citationsarXiv
Transformers Provably Learn to Internalize Chain-of-Thought2605.28600v1 · Yixiao Huang, Hanlin Zhu, Zixuan Wang et al.2026 · 0 citationsarXiv
MLS-Bench: A Holistic and Rigorous Assessment of AI Systems on Building Better AI2605.08678v3 · Bohan Lyu, Yucheng Yang, Siqiao Huang et al.2026 · 0 citationsarXiv
PivotRL: High Accuracy Agentic Post-Training at Low Compute Cost2603.21383v1 · Junkeun Yi, Damon Mosk-Aoyama, Baihe Huang et al.2026 · 0 citationsarXiv
Towards Anytime-Valid Statistical Watermarking2602.17608v1 · Baihe Huang, Eric Xu, Kannan Ramchandran et al.2026 · 0 citationsarXiv
dUltra: Ultra-Fast Diffusion Language Models via Reinforcement Learning2512.21446v2 · Shirui Chen, Jiantao Jiao, Lillian J. Ratliff et al.2025 · 0 citationsarXiv
From Bits to Rounds: Parallel Decoding with Exploration for Diffusion Language Models2511.21103v1 · Hengyu Fu, Baihe Huang, Virginia Adams et al.2025 · 0 citationsarXiv
Emergence of Superposition: Unveiling the Training Dynamics of Chain of Continuous Thought2509.23365v3 · Hanlin Zhu, Shibo Hao, Zhiting Hu et al.2025 · 0 citationsarXiv
GSM-Agent: Understanding Agentic Reasoning Using Controllable Environments2509.21998v2 · Hanlin Zhu, Tianyu Guo, Song Mei et al.2025 · 0 citationsarXiv
Generalization or Hallucination? Understanding Out-of-Context Reasoning in Transformers2506.10887v4 · Yixiao Huang, Hanlin Zhu, Tianyu Guo et al.2025 · 0 citationsarXiv
Sample Complexity and Representation Ability of Test-time Scaling Paradigms2506.05295v2 · Baihe Huang, Shanda Li, Tianhao Wu et al.2025 · 0 citationsarXiv
Reasoning by Superposition: A Theoretical Perspective on Chain of Continuous Thought2505.12514v3 · Hanlin Zhu, Shibo Hao, Zhiting Hu et al.2025 · 0 citationsarXiv
How Do LLMs Perform Two-Hop Reasoning in Context?2502.13913v2 · Tianyu Guo, Hanlin Zhu, Ruiqi Zhang et al.2025 · 0 citationsarXiv
Token Assorted: Mixing Latent and Text Tokens for Improved Language Model Reasoning2502.03275v2 · DiJia Su, Hanlin Zhu, Yingchen Xu et al.2025 · 1 citationarXiv
How to Evaluate Reward Models for RLHF2410.14872v2 · Evan Frick, Tianle Li, Connor Chen et al.2024 · 0 citationsarXiv
Active-Dormant Attention Heads: Mechanistically Demystifying Extreme-Token Phenomena in LLMs2410.13835v2 · Tianyu Guo, Druv Pai, Yu Bai et al.2024 · 0 citationsarXiv
Thinking LLMs: General Instruction Following with Thought Generation2410.10630v1 · Tianhao Wu, Janice Lan, Weizhe Yuan et al.2024 · 3 citationsarXiv
EmbedLLM: Learning Compact Representations of Large Language Models2410.02223v2 · Richard Zhuang, Tianhao Wu, Zhaojin Wen et al.2024 · 0 citationsarXiv
Meta-Rewarding Language Models: Self-Improving Alignment with LLM-as-a-Meta-Judge2407.19594v2 · Tianhao Wu, Weizhe Yuan, Olga Golovneva et al.2024 · 12 citationsarXiv
Information-driven design of imaging systems2405.20559v5 · Henry Pinkard, Leyla Kabuli, Eric Markley et al.2024 · 2 citationsarXiv
Toxicity Detection for Free2405.18822v2 · Zhanhao Hu, Julien Piet, Geng Zhao et al.2024 · 3 citationsarXiv
Towards a Theoretical Understanding of the 'Reversal Curse' via Training Dynamics2405.04669v2 · Hanlin Zhu, Baihe Huang, Shaolun Zhang et al.2024 · 1 citationarXiv
Toward a Theory of Tokenization in LLMs2404.08335v2 · Nived Rajaraman, Jiantao Jiao, Kannan Ramchandran2024 · 4 citationsarXiv
Generative AI Security: Challenges and Countermeasures2402.12617v2 · Banghua Zhu, Norman Mu, Jiantao Jiao et al.2024 · 9 citationsarXiv
Efficient Prompt Caching via Embedding Similarity2402.01173v1 · Hanlin Zhu, Banghua Zhu, Jiantao Jiao2024 · 1 citationarXiv
Iterative Data Smoothing: Mitigating Reward Overfitting and Overoptimization in RLHF2401.16335v1 · Banghua Zhu, Michael I. Jordan, Jiantao Jiao2024 · 0 citationsarXiv
Towards Optimal Statistical Watermarking2312.07930v3 · Baihe Huang, Hanlin Zhu, Banghua Zhu et al.2023 · 1 citationarXiv
End-to-end Story Plot Generator2310.08796v1 · Hanlin Zhu, Andrew Cohen, Danqing Wang et al.2023 · 0 citationsarXiv
Pairwise Proximal Policy Optimization: Harnessing Relative Feedback for LLM Alignment2310.00212v3 · Tianhao Wu, Banghua Zhu, Ruoyu Zhang et al.2023 · 4 citationsarXiv
Guided Online Distillation: Promoting Safe Reinforcement Learning by Offline Demonstration2309.09408v2 · Jinning Li, Xinyi Liu, Banghua Zhu et al.2023 · 5 citationsarXiv
Noisy Computing of the and Functions2309.03986v1 · Banghua Zhu, Ziao Wang, Nadim Ghaddar et al.2023 · 0 citationsarXiv
On the Optimal Bounds for Noisy Computing2306.11951v1 · Banghua Zhu, Ziao Wang, Nadim Ghaddar et al.2023 · 0 citationsarXiv
Fine-Tuning Language Models with Advantage-Induced Policy Alignment2306.02231v3 · Banghua Zhu, Hiteshi Sharma, Felipe Vieira Frujeri et al.2023 · 6 citationsarXiv
On Optimal Caching and Model Multiplexing for Large Model Inference2306.02003v2 · Banghua Zhu, Ying Sheng, Lianmin Zheng et al.2023 · 3 citationsarXiv
Doubly Robust Self-Training2306.00265v3 · Banghua Zhu, Mingyu Ding, Philip Jacobson et al.2023 · 0 citationsarXiv
Online Learning in a Creator Economy2305.11381v1 · Banghua Zhu, Sai Praneeth Karimireddy, Jiantao Jiao et al.2023 · 4 citationsarXiv
Statistical Complexity and Optimal Algorithms for Non-linear Ridge Bandits2302.06025v3 · Nived Rajaraman, Yanjun Han, Jiantao Jiao et al.2023 · 1 citationarXiv
Importance Weighted Actor-Critic for Optimal Conservative Offline Reinforcement Learning2301.12714v2 · Hanlin Zhu, Paria Rashidinejad, Jiantao Jiao2023 · 5 citationsarXiv
Online Learning in Stackelberg Games with an Omniscient Follower2301.11518v2 · Geng Zhao, Banghua Zhu, Jiantao Jiao et al.2023 · 2 citationsarXiv
Principled Reinforcement Learning with Human Feedback from Pairwise or -wise Comparisons2301.11270v5 · Banghua Zhu, Jiantao Jiao, Michael I. Jordan2023 · 21 citationsarXiv
The Sample Complexity of Online Contract Design2211.05732v3 · Banghua Zhu, Stephen Bates, Zhuoran Yang et al.2022 · 0 citationsarXiv
Optimal Conservative Offline RL with General Function Approximation via Augmented Lagrangian2211.00716v1 · Paria Rashidinejad, Hanlin Zhu, Kunhe Yang et al.2022 · 0 citationsarXiv
Career total: 141 works. 100 are in this corpus.Showing the 50 most recent.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.