TraFL applies trajectory flow balancing to post-train diffusion language models, preventing mode collapse and delivering consistent gains on reasoning tasks that hold under increased sampling.
Livecodebench: Holistic and contamination free evaluation of large language models for code
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
citation-role summary
background 1
dataset 1
citation-polarity summary
representative citing papers
Diversity collapse in LLMs arises from order and shape miscalibration in token probability distributions at inference time, not from sampling methods.
Empirical study shows LLM inference backends can shift benchmark scores by up to 16.6 percentage points and cause output disagreements due to optimizations like prefix caching and custom kernels.
LAQuant improves long-decoding accuracy on quantized reasoning models like Qwen3-4B by 15pp on AIME25 via layer-wise lookahead loss, achieving 3.42x speedup over FP16.
MTraining scales LLM training to 512K-token contexts on 32 A100 GPUs by integrating dynamic sparse training patterns with balanced and hierarchical sparse ring attention, achieving up to 6x throughput gains without accuracy loss on long-context benchmarks.
Systematic study of inter-agent communication in LLM multi-agent systems shows reasoning and verification are critical for performance, with a new augmentation technique recovering 86.2% of failures.
SnapMLA achieves up to 1.91x higher throughput in long-output MLA decoding using FP8 quantization and specialized kernels while keeping benchmark quality near the BF16 baseline.
citing papers explorer
-
Beyond Mode-Seeking RL: Trajectory-Balance Post-Training for Diffusion Language Models
TraFL applies trajectory flow balancing to post-train diffusion language models, preventing mode collapse and delivering consistent gains on reasoning tasks that hold under increased sampling.
-
Sampling More, Getting Less: Calibration is the Diversity Bottleneck in LLMs
Diversity collapse in LLMs arises from order and shape miscalibration in token probability distributions at inference time, not from sampling methods.
-
The Silent Hyperparameter: Quantifying the Impact of Inference Backends on LLM Reproducibility
Empirical study shows LLM inference backends can shift benchmark scores by up to 16.6 percentage points and cause output disagreements due to optimizations like prefix caching and custom kernels.
-
LAQuant: A Simple Overhead-free Large Reasoning Model Quantization by Layer-wise Lookahead Loss
LAQuant improves long-decoding accuracy on quantized reasoning models like Qwen3-4B by 15pp on AIME25 via layer-wise lookahead loss, achieving 3.42x speedup over FP16.
-
MTraining: Distributed Dynamic Sparse Attention for Efficient Ultra-Long Context Training
MTraining scales LLM training to 512K-token contexts on 32 A100 GPUs by integrating dynamic sparse training patterns with balanced and hierarchical sparse ring attention, achieving up to 6x throughput gains without accuracy loss on long-context benchmarks.
-
What Do Agents Communicate? Characterizing Information Exchange in Multi-Agent Systems
Systematic study of inter-agent communication in LLM multi-agent systems shows reasoning and verification are critical for performance, with a new augmentation technique recovering 86.2% of failures.
-
SnapMLA: Efficient Long-Context MLA Decoding via Hardware-Aware FP8 Quantized Pipelining
SnapMLA achieves up to 1.91x higher throughput in long-output MLA decoding using FP8 quantization and specialized kernels while keeping benchmark quality near the BF16 baseline.