Category
8,685 papers
The introduction of WhiteMatter, which allows every layer to draw on past-token representations from any depth, and cyclic iteration, which updates interleaved groups of tokens in turn while processing the tokens within each group in parallel.
We use training-data attribution as an interpretable tool for capability discovery, mapping which regions of the pretraining corpus support social reasoning versus STEM reasoning in OLMo3-7B. Training-data attribution measures how strongly each training document influences a model's predictions on a benchmark, but docu...
Experiments show that Reflective On-policy Self-Distillation (ROSD), which distills transferable reasoning insights rather than complete reference trajectories, consistently outperforms standard OPSD and reinforcement learning baselines, better preserves reasoning diversity, stabilizes training, and mitigates late-stag...
This work revisits classical sentence classification models from a margin enlargement perspective, aggregating token-level features via max pooling and directly estimating sentence scores using a lightweight MLP, and achieves substantial efficiency improvements while maintaining competitive performance with state-of-th...
Reach audiences
Advertise in front of researchers, engineers, and readers.
Speech representations from self-supervised speech models (S3Ms) are known to be sensitive to phonemic contrasts, but their sensitivity to prosodic contrasts has not been directly measured. The ABX discrimination task has been used to measure phonemic contrast in S3M representations via minimal pairs. We introduce pros...
This work introduces Fast-Slow Thinking Reward Models (F/S-RM), a hybrid RM architecture inspired by Dual Process Theory that achieves state-of-the-art accuracy with an average score of 84.3 across benchmarks, while reducing token consumption by 22.5%.
Large language models (LLMs) exhibit strong performance on self-contained programming tasks. However, they still struggle with repository-level software engineering (SWE), which demands (1) deep codebase navigation with effective context management for accurate localization, and (2) systematic approaches for iterative,...
Rule-to-Tag (R2T), a framework that turns linguistic rules into the training signal for neural sequence taggers in low-resource languages, is introduced, a first instance of a broader paradigm the authors call principled learning (PrL): using explicit principles as a learning signal, rather than relying on example-base...
Choosing a Large Language Model (LLM) for a given task requires comparing many strong candidates, yet standard evaluation relies on costly annotations over fixed evaluation sets. To address this challenge, we develop SELECT-LLM, the first framework for active model selection of LLMs. SELECT-LLM aims to find a small set...
LaSEr-Edit is proposed, a constraint-satisfying text revision method that can be applied to any LLMs, black- or white-box, and it is found that lightweight, task-specific energy-based models (EBMs) achieve error-localization performance competitive with or even better than that of much larger LLMs, while operating subs...
Granularity alignment, not algorithm choice, localizes when routing helps, and what category supervision buys when deployed through routing is quantified.
EpiKV is introduced, which combines hidden-state shifts with the model's recent query--key relevance to rank cached tokens without materializing the attention matrix, and matches or outperforms the strongest attention-based eviction baselines while running directly in vLLM with unmodified attention kernels.
Flint-interactive gives AI agents a language for creating interactive visualizations. Click a chart, filter data, highlight related content, and explore beyond the view the author originally chose. The post Flint-interactive: A language for creating interactive data visualizations appeared first on Microsoft Research.
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
A new computational approach reveals how cells communicate to assemble into tissues. This could accelerate stem cell engineering, improve disease models, and inform therapies.
Zero-knowledge proofs can verify computing results without rerunning workloads or exposing sensitive data. Magnetar helps bring them into the mainstream with support for standard programming languages, major speed gains, and quantum-resistant security. The post Proof-based general-purpose computing with Magnetar appeared first on Microsoft Research.
A new language-processing tool could help identify the highest-risk individuals from natural language, enabling swifter interventions.
A Blog post by Liquid AI on Hugging Face
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
