paper

Self-Augmenting Retrieval for Diffusion Language Models

paperactiveprovisionalself-augmenting-retrieval-for-diffusion-language-models-b1d03fa5·1 events·first seen 12d ago

Aliases: Self-Augmenting Retrieval for Diffusion Language Models

Co-occurring entities

More like this (12)

Diffusion Language Models LESS: Mutual-Stability Sampling for Diffusion Language Models Knowledge Editing in Masked Diffusion Language Models Beyond Fully Random Masking: Attention-Guided Denoising and Optimization for Diffusion Language Models continuous diffusion language model diffusion-based generative models Representation-Conditioned Diffusion Models When Does Mixing Help? Analyzing Query Embedding Interpolation in Multilingual Dense Retrieval Masked Diffusion Models A Diffusion Approximation for Temporal-Difference Learning with Linear Features under Markovian Noise Denoising Diffusion Probabilistic Models Reinforcement Learning for Language Models

Recent events (1)

5arXiv · cs.LG·12d ago·source ↗

SARDI: Self-Augmenting Retrieval for Diffusion Language Models using lookahead tokens

Researchers introduce SARDI, a training-free RAG framework for discrete diffusion language models that repurposes discarded low-confidence tokens during denoising as lookahead signals to guide retrieval before output is finalized. The method is retriever-agnostic and applicable to any reasoning-capable discrete diffusion LM. Evaluated across five multi-hop QA benchmarks, SARDI outperforms training-free diffusion and autoregressive retrieval baselines at up to 8x higher throughput.

Evaluation and Benchmarking Agent and Tool Ecosystem Self-Augmenting Retrieval for Diffusion Language Models SARDI