Category
2,863 papers
Understanding gender biases in large language models (LLMs) is increasingly important as these systems become embedded in decision-support tools with real consequences. Prior research has focused only on a small set of models, leaving open the extent to which gender biases are common and heterogeneous across LLMs. We a...
FAIR PERSONALIZATION & EXTERNAL VALIDITY AT THE LIMITCalibration Privilege, Exclusion Risk, Subgroup Uncertainty, Cross-Site Transfer, Device Heterogeneity, and Equitable Adaptive BCI Feng Cheng-en (33) x Starli Does personalization help people differently - or does it simply make access easier for the people who were...
This Zenodo record is a permanently preserved version of a PREreview. You can view the complete PREreview at https://prereview.org/reviews/23071859. Authors of the review Name: Stephanie R. U, ORCID: https://orcid.org/0000-0002-6220-4661 Bio: Stephanie U is a PhD candidate whose work investigates the role of circadian...
This work examines both model behavior and internal representations, using the empirical neural tangent kernel (NTK) as the primary diagnostic tool, and develops practical guidance for designing DR schemes, selecting models, and detecting shortcut learning.
Reach audiences
Advertise in front of researchers, engineers, and readers.
End-to-end runtime analysis further confirms that the overhead introduced by tracking, CLIP-based recurrence matching, and the identifiability gate remains modest relative to the dominant Sa2VA inference cost, thereby validating the efficiency of the proposed pipeline.
NAMOH, an architecture-native sparse attention mechanism that activates only its assigned tokens and performs causal attention within this subsequence, is introduced, and it is hoped this work offers a new path for scaling attention, with parameter scaling directly enabling context scaling.
Web text makes up the majority of pretraining data and is increasingly AI-generated. After applying FineWeb quality filtering, we find that 27.5% of tokens from June 2026 web data are labeled as AI-generated by Pangram, rising to 31.1% by August. Unlike synthetic data or model-collapse setups, this *wild* AI text comes...
Recent advances in large language models have made automatic game generation increasingly feasible, yet reliably improving generated games beyond a playable version remains challenging. Naive iterative refinement can easily overfit a small set of test cases, producing fragile games with unresolved bugs, missing behavio...
Representing scientific papers as points in a space lets us search for similar papers and inquire about how fields relate to one another and drive innovation. Beyond search, the vector space of papers invites generation: mixing papers through simple vector operations creates new points, mirroring combinatorial novelty,...
Black-box distillation is a practical route for transferring capabilities from API-accessible large language models that expose only text outputs into smaller student models. Recent on-policy adversarial methods such as GAD improve over SeqKD by forming an adversarial loop between a critic and a student, where the crit...
This study systematically evaluates self-evaluation-based uncertainty quantification across different language models of varying sizes on question-answering tasks spanning general to specialized knowledge domains. Using various self-evaluation methods where models judge their own predictions, we examine how model scale...
Causal interventions such as activation patching and distributed alignment search (DAS) are the main tool for making mechanistic claims about neural networks. Recent work showed that these interventions routinely push representations off the model's natural distribution, and that such divergence is sometimes harmless a...
Earth AI
A Blog post by Ai2 on Hugging Face
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
