arXiv Machine Learning October 22, 2026 By Zaynab Merimi, Noussayba El Marrakchi, Arthur Satouf

A Diagnostic Study of Faithfulness in Diffusion Language Models for Table-to-Text Generation

Read the original on arXiv Machine Learning →

The Flow has not summarised this story yet — read it at arXiv Machine Learning.

arXiv Machine Learning

Jun 5

Self-Augmenting Retrieval for Diffusion Language Models

arXiv:2606. 06474v1 Announce Type: cross Abstract: Discrete diffusion language models generate text by iteratively denoising an entire response in parallel.

By Paul J\"unger, Justin Lovelace, Linxi Zhao, Dongyoung Go, Kilian Q. Weinberger

llms rag diffusion efficiency benchmarks

arXiv Computation and Language

2d ago

Hacking Generative Perplexity: Why Unconditional Text Evaluation Needs New Metrics

The paper critiques generative perplexity (gen‑PPL) as an unreliable metric for evaluating unconditional text generation, showing that it can be minimized by trivial, non‑diverse outputs. The authors demonstrate this by constructing simple samplers that achieve low gen‑PPL yet produce incoherent text, and they propose alternative evaluation tools—including distributional discrepancies, diversity diagnostics, and a compression‑based KL divergence measure—to better assess generative language models. They argue that future progress in diffusion language models should rely on a suite of metrics that capture diverse aspects of generated text quality.

By Antonio Franca, Alexander Tong

llms diffusion

arXiv AI

Jul 14

Prompt Compression in Diffusion Large Language Models: Evaluating LLMLingua-2 on LLaDA

arXiv:2605. 17932v2 Announce Type: replace-cross Abstract: Prompt compression reduces inference cost and context length in large language models, but prior evaluations focus mainly on autoregressive architectures.

By Sterling Huang, Abigayle Brown, Jiyoo Noh, Jiakang Xu, Wantong Huo, Kaung Myat Kyaw, Jonathan Chan

llms diffusion nlp

arXiv AI

Jun 9

Hacking Generative Perplexity: Why Unconditional Text Evaluation Needs Distributional Metrics

arXiv:2606. 08417v1 Announce Type: cross Abstract: Diffusion and continuous flow-based language models have emerged as the leading non-autoregressive alternatives to language modeling.

By Antonio Franca, Alexander Tong

llms diffusion benchmarks safety

arXiv Computation and Language

4d ago

The Bottleneck Is Retrieval, Not Reading: Diagnosing Evidence-Based Generation with RAG for Long Context Question Answering

arXiv:2610.26951v1 Announce Type: new Abstract: Retrieval-augmented generation (RAG) is the dominant paradigm for grounding large language models in external corpora, but its similarity-driven, chunk...

By Debopriyo Banerjee, Owais Aijaz, Ahmer Tabassum, Zhuohan Xie, Preslav Nakov, Veselin Stoyanov

llms rag nlp benchmarks

Hugging Face Trending Papers

Oct 6

Disentangling Paradigm, Identifier, and Decoding in Generative Retrieval

Generative retrieval trains a language model to generate the identifier of a relevant document. Recent work replaces the autoregressive decoder with diffusion, but changes identifiers, training recipe...