Category
3,366 papers
It is argued that benchmarks should stratify on articulation, and the extracted structure is offered as a measurement instrument for doing so.
It is shown that the vector representations of a variety of neural networks can be closely approximated with symbolic structures, providing a potential way to reconcile longstanding symbolic conceptions of intelligence with the vector-based nature of modern AI.
Engram Adapter, a framework that repurposes pretraining-time conditional memory as a post-hoc adapter for frozen LLMs, improves in-domain accuracy while preserving 99.4%--100.1% of average OOD performance; on LegalBench it slightly exceeds the frozen base model on average, whereas comparable always-on baselines degrade...
This work reports on building and operating a retrieval-augmented legal assistant for Uzbek that must run in two regimes: a managed cloud service that maximizes answer quality within a per-token cost ceiling, and an on-premises deployment for clients whose legal data may not leave their infrastructure, restricting us t...
Reach audiences
Advertise in front of researchers, engineers, and readers.
A lightweight inference-time correction that estimates and removes a popularity prior from model predictions is proposed, which requires no fine-tuning, is label-free at test time, and adds negligible computational cost.
RouteSparse is introduced, which routes each head and prompt segment among a small library of GPU-efficient sparse patterns, andlations confirm that input-conditional routing, hardware profiling, and selective dense fallback each contribute to the quality--latency tradeoff.
This work proposes using Tensor Product Representations (TPRs) as a unifying hypothesis, and shows that TPRs can unify several prior interpretability methods: additive analogies, linear probing, sparse autoencoders, and activation patching.
The results indicate that current MLLMs can process visually clear English and structured numeric content, but reliable native Khmer document understanding remains an open challenge.
TopoCompress is introduced, a training-free and model-agnostic framework that compresses long contexts by selecting coherent semantic spans by selecting coherent semantic spans and achieves performance comparable to the strongest baseline while using a 4x smaller compression budget.
This work compares a ~28M-parameter autoregressive Tiny Recursive Model (TRM-AR) on natural-language-to-Python code generation against parameter-matched and depth-matched controls, tracking fit and generation across 40 epochs and three seeds.
A proof-of-concept Brain-Language-Action (BLA) model for drone control using motor-imagery EEG from the BCI Competition IV 2a dataset is developed, providing an initial demonstration that language conditioning can expand the effective control range of EEG-based robotic interfaces without requiring a corresponding incre...
This work validates the engineering feasibility of running industrial-scale trillion-parameter LLM-driven biomedical computing tasks on consumer hardware, establishing a new low-barrier paradigm for AI-powered early stage drug discovery.
Generative AI
Assistant Professor Benjamin Lindquist studies how engineers deliberately injected randomness into early AI, giving rise to deep-learning models that sometimes hallucinate.
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
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.
A Blog post by IBM Research on Hugging Face
A Blog post by IBM Research on Hugging Face
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
