7 papers
APEX4: Efficient Pure W4A4 LLM Inference via Intra-SM Compute Rebalancing
Hong Guo, Nianhui Guo, Weixing Wang +3
W4A4 quantization promises full utilization of INT4 Tensor Cores, yet group dequantization overhead on CUDA Cores has driven existing systems to mixed-precision fallbacks. We prese…
Sample Where You Struggle: Sharpening Base Model Reasoning via Entropy-Guided Power Sampling
Hong Guo, Nianhui Guo, Christoph Meinel +1
Sampling from the sequence-level power distribution elicits RL-level reasoning from base language models without any parameter updates, but the standard Metropolis--Hastings…
Harmonizing Dense and Sparse Signals in Multi-turn RL: Dual-Horizon Credit Assignment for Industrial Sales Agents
Haojin Yang, Ai Jian, Xinyue Huang +5
Optimizing large language models for industrial sales requires balancing long-term commercial objectives (e.g., conversion rate) with immediate linguistic constraints such as fluen…
WavefrontDiffusion: Dynamic Decoding Schedule for Improved Reasoning
Haojin Yang, Rui Hu, Zequn Sun +3
Diffusion Language Models (DLMs) have shown strong potential for text generation and are becoming a competitive alternative to autoregressive models. The denoising strategy plays a…
VADE: Variance-Aware Dynamic Sampling via Online Sample-Level Difficulty Estimation for Multimodal RL
Zengjie Hu, Jiantao Qiu, Tianyi Bai +5
Group-based policy optimization methods like GRPO and GSPO have become standard for training multimodal models, leveraging group-wise rollouts and relative advantage estimation. Ho…
Image Tokens Matter: Mitigating Hallucination in Discrete Tokenizer-based Large Vision-Language Models via Latent Editing
Weixing Wang, Zifeng Ding, Jindong Gu +4
Large Vision-Language Models (LVLMs) with discrete image tokenizers unify multimodal representations by encoding visual inputs into a finite set of tokens. Despite their effectiven…