Category
9,568 papers
This work develops a conception of understanding fit for polyphonic AI that centres on sound circuitry that is reliably and correctly recruited and in control of outputs, and can do the work of guiding trust in AI.
This work quantifies the compression cost of boundaries while separating it from the prediction quality of the resulting token units, and introduces boundary licences, which limit the vocabulary entries permitted to cross cuts and admit the same form of certificate.
This work presents a neurosymbolic router that classifies each incoming query and dispatches it to the cheapest correct solver, sending structured tasks to deterministic engines and reserving the small language model (SLM) for open-ended word problems.
A Wizard-of-Oz study design and supporting real-time system for collecting multimodal behavioral data from native Chinese speakers reacting to social norm violations by German learners.
Reach audiences
Advertise in front of researchers, engineers, and readers.
This paper presents AuditBench, a new benchmark dataset for evaluating the capabilities of LLMs at investigating security-related system audit logs. We design and use this benchmark to explore the performance of LLMs on four log-investigation tasks that incident response teams commonly perform, ranging from triaging al...
This work formalizes agent attribution as a cooperative game, parameterized by the coalition distribution, intervention protocol, and target metric, which provides a principled approach for cost-effective MAS attribution and intervention.
The rapid growth of online video platforms and AI-generated content has made reliable video guardrails a key challenge for safety and real-world deployment. While most videos can be screened through fast pattern recognition, a small subset requires deeper reasoning over temporally complex content and nuanced policy con...
As large language models (LLMs) are increasingly considered for automated assessment and feedback, understanding when LLM marking is valid is essential. We evaluate LLM-as-a-judge marking across four settings spanning three assessment formats - structured questions, written essays, and scientific plots - comparing GPT-...
ChEmbed is introduced, the first purpose-built family of domain-adapted text embedding models engineered for chemical literature retrieval, and represents a practical, lightweight, and reproducible embedding solution that effectively improves chemical literature retrieval.
The reachability problem in symmetric vector addition systems with states (VASS), where transitions are invariant under a group of permutations of coordinates, is investigated and the gain in complexity when the group arises as a combination of the trivial and symmetric groups is estimated.
D-STT, a simple yet effective defense algorithm that identifies and explicitly decodes safety trigger tokens of the given safety-aligned LLM to activate the model's learned safety patterns, effectively preserves model usability by introducing minimum intervention in the decoding process.
The proposed Uncertainty-Calibrated MOPD (UCMOPD), which addresses limitations of standard MOPD through two complementary mechanisms, improves the general-capability average over standard MOPD by $4.48\% and $7.86\% while maintaining vertical-domain performance.
Natural Language Processing
Our most capable vision-language-action model enhances robot dexterity, reliability, and safety
A Blog post by Ai2 on Hugging Face
A Blog post by IBM Research on Hugging Face
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.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
