16 papers
Self-Evolving Agents through Privileged Retrospective Distillation
Yuhang Zhou, Yifan Wu, Bo Peng +5
Self-evolving agents aim to improve from their own experience rather than remain fixed after deployment, but naturally collected trajectories often lack ground-truth solutions or v…
Verify Less, Evolve More: Training Idea-Level Critics for Verification-Efficient ML Evolving Agents
Jiamu Bai, Lizhu Zhang, Xin Yu +9
As large language models become more powerful, self-evolving agents are able to tackle challenging tasks including AI for machine learning (AI4ML). In AI4ML, while empirical verifi…
Discovered, Not Designed: Population Evolution for Collaborative and Compute-Intensive Model Discovery
Bo Peng, Lizhu Zhang, Yuhang Zhou +5
LLM-driven evolution enables iterative model development, but two practical goals remain underexplored: finding model designs that transfer across related tasks and sustaining impr…
Mixture of Self-Improving Branches For Agent Harness Optimization
Haoyu Dong, Yuhang Zhou, Zihao Lin +6
Harness optimization provides a practical setting for recursive self-improvement (RSI), where agent-generated modifications inform subsequent changes through execution feedback. Re…
MLToolBench: Learning Tool-Augmented Agents for Machine Learning Development
Xin Yu, Lizhu Zhang, Jiamu Bai +7
Machine learning engineering (MLE) agents have made substantial progress, but learning through ML experimentation remains costly in time and computation. Synthetic environments red…
Remember When It Matters: Proactive Memory Agent for Long-Horizon Agents
Yifan Wu, Lizhu Zhang, Yuhang Zhou +5
In long-horizon tasks, decision-relevant state is often scattered across an expanding trajectory, while the action agent must surface it and act. As trajectories grow, task require…
SAGE-OPD: Selective Agent-Guided Intervention for Multi-Turn On-Policy Distillation
Yuhang Zhou, Lizhu Zhang, Yifan Wu +5
On-policy distillation (OPD) improves student models by training them on trajectories induced by their own policy, making it a promising approach for mitigating exposure bias in ag…
OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification
Yuhang Zhou, Lizhu Zhang, Yifan Wu +5
On-Policy Distillation (OPD) trains a student model on its own generative trajectories under dense token-level feedback from a stronger teacher, mitigating both the off-policy dist…
Multi-Token Residual Prediction
Yufeng Xu, Zishuo Bao, Qian Wang +6
Diffusion Language Models (DLMs) generate text by iteratively denoising masked token sequences, offering a tradeoff between parallelism and quality compared to autoregressive model…
Efficient Pre-Training with Token Superposition
Bowen Peng, Théo Gigant, Jeffrey Quesnelle
Pre-training of Large Language Models is often prohibitively expensive and inefficient at scale, requiring complex and invasive modifications in order to achieve high data throughp…
Decoupling the Benefits of Subword Tokenization for Language Model Training via Byte-level Simulation
Théo Gigant, Bowen Peng, Jeffrey Quesnelle
Subword tokenization is an essential part of modern large language models (LLMs), yet its specific contributions to training efficiency and model performance remain poorly understo…
Topology-Enhanced Alignment for Large Language Models: Trajectory Topology Loss and Topological Preference Optimization
Yurui Pan, Ke Xu, Bo Peng
Alignment of large language models (LLMs) via SFT and RLHF/DPO typically ignores the global geometry of the representation space, relying instead on local token likelihoods or scal…
PACEvolve++: Improving Test-time Learning for Evolutionary Search Agents
Minghao Yan, Bo Peng, Benjamin Coleman +11
Large language models have become drivers of evolutionary search, but most systems rely on a fixed, prompt-elicited policy to sample next candidates. This limits adaptation in prac…
Long Context Pre-Training with Lighthouse Attention
Bowen Peng, Subho Ghosh, Jeffrey Quesnelle
Training causal transformers at extreme sequence lengths is bottlenecked by the quadratic time and memory of scaled dot-product attention (SDPA). In this work, we propose Lighthous…
AgenticTagger: Structured Item Representation for Recommendation with LLM Agents
Zhouhang Xie, Bo Peng, Zhankui He +11
High-quality representations are a core requirement for effective recommendation. In this work, we study the problem of LLM-based descriptor generation, i.e., keyphrase-like natura…
PACEvolve: Enabling Progress-Aware Consistent Evolution
Minghao Yan, Bo Peng, Benjamin Coleman +13
Self-evolving agents powered by Large Language Models (LLMs) have emerged as a promising direction across diverse domains, including code optimization and scientific discovery, yet…