SOL: Measuring Gaps between Text Distributions by Double Sliced Wasserstein Metrics
cs.CL, cs.LG, stat.ML
Submitted: 2026-10-05
Updated: 2026-10-08
Code: https://github.com/openai/gpt-2-output-dataset
Sources
- Block Diffusion: Interpolating Between Autoregressive and Diffusion Language Models
- MIND: Monge Inception Distance for Generative Models Evaluation
- NeoBERT: A Next-Generation BERT
- Exploring Precision and Recall to assess the quality and diversity of LLMs
- One Billion Word Benchmark for Measuring Progress in Statistical Language Modeling
- Spherical Flows for Sampling Categorical Data
- LangFlow: Continuous Diffusion Rivals Discrete in Language Modeling
- Continuous diffusion for categorical data
- Hacking Generative Perplexity: Why Unconditional Text Evaluation Needs Distributional Metrics
- Uniform Diffusion Models Revisited: Leave-One-Out Denoiser and Absorbing State Reformulation
- Likelihood-Based Diffusion Language Models
- Diffusion-LM Improves Controllable Text Generation
- Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution
- 2 OLMo 2 Furious
- Generalized Wasserstein Flow Matching: Transport Plans, Everywhere, All at Once
- Simple and Effective Masked Diffusion Language Models
- Text Embeddings by Weakly-Supervised Contrastive Pre-training
- Qwen3 Technical Report
- Dream 7B: Diffusion Large Language Models
- Self-conditioned Flow Map Language Models via Fixed-point Flows
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering