8 papers
Self-Improving Large Language Models via Progressive Experience Evolution
Shijie Ren, Xiting Wang, Meng Li +8
Large language models (LLMs) capable of self-improvement require not only effective policy optimization, but also a principled mechanism for transforming transient interaction expe…
LEEPS: Latent-Guided Explore-Exploit Prompt Sampling for Efficient RLVR in Large Language Models
Shuang Liang, Haoyang Zhou, Yifan Gong +2
Reinforcement learning with verifiable rewards (RLVR) improves the reasoning capabilities of large language models, but prompt groups with identical rollout rewards consume generat…
NKI-Agent: Domain-Specific Fine-Tuning and Agentic Tool Use for Neuron Kernel Generation
Junjie Tang, Jun Huan, Hao Zhou +2
Recent agentic approaches to LLM-based kernel generation have achieved impressive results on CUDA. For emerging AI accelerators such as AWS Trainium and Inferentia, automated kerne…
TAD: Temporal-Aware Trajectory Self-Distillation for Fast and Accurate Diffusion LLM
Haoyang Zhou, Li Kong, Shijie Ren +4
Diffusion large language models (dLLMs) offer a promising paradigm for parallel text generation, but in practice they face an accuracy-parallelism trade-off, where increasing token…
Safety Anchor: Defending Harmful Fine-tuning via Geometric Bottlenecks
Guoxin Lu, Letian Sha, Qing Wang +4
The safety alignment of Large Language Models (LLMs) remains vulnerable to Harmful Fine-tuning (HFT). While existing defenses impose constraints on parameters, gradients, or intern…
Collage: Light-Weight Low-Precision Strategy for LLM Training
Tao Yu, Gaurav Gupta, Karthick Gopalswamy +7
Large models training is plagued by the intense compute cost and limited hardware memory. A practical solution is low-precision representation but is troubled by loss in numerical…
Variance-reduced Zeroth-Order Methods for Fine-Tuning Language Models
Tanmay Gautam, Youngsuk Park, Hao Zhou +2
Fine-tuning language models (LMs) has demonstrated success in a wide array of downstream tasks. However, as LMs are scaled up, the memory requirements for backpropagation become pr…
HLAT: High-quality Large Language Model Pre-trained on AWS Trainium
Haozheng Fan, Hao Zhou, Guangtai Huang +6
Getting large language models (LLMs) to perform well on the downstream tasks requires pre-training over trillions of tokens. This typically demands a large number of powerful compu…