Category
3,629 papers
SinkProbe is built, a suite that measures sink mass, massive activation, position resolved recall and the recency gap, and applies it to four small models that differ only in how they mix tokens and depth, and three results follow.
The industrial process monitoring foundation model (IPM-FM) first learns general-purpose representations from unlabeled industrial process data through self-supervised pretraining, then adapts to specific monitoring tasks using a small amount of task-labeled data, and finally produces calibrated predictions through an...
A*-Thought-V2 is presented, a geometric dynamics of LLM guided framework that models CoT as a hidden-state trajectory and replaces hard deletion with an explicit-implicit interleaved latent architecture that reflects broader soft targets that encourage richer step-level feature learning.
This work presents Noesis, the deterministic-first query plane of the Noesis architecture, which makes every deterministic judgment before generation, and describes its mechanisms, which follow from the ingestion architecture.
Reach audiences
Advertise in front of researchers, engineers, and readers.
LoGAN, a VLM-based agentic framework for few-shot multilingual font localization, which takes in a small number of individual glyphs from a font or letters from a logo and uses them to generate complete character sets in other languages, achieves broad language coverage for font localization with various styles.
Activation steering promises modular control of language model behavior: a behavior such as politeness corresponds to a direction in a model's activations, and adding that direction while it generates should switch the behavior on and leave everything else alone. It does not. We ask what decides which other behaviors m...
Models trained from scratch with safety co-training spread continuously across pretraining reach 87 to 98% AdvBench refusal that the same attack erodes by only 2 to 14 pp at every scale from 410M to 6.9B, at a small cost on short-answer capability probes.
The analysis shows that CoTs do not reliably track visual evidence that influences model predictions, and it is found that Predict-then-Explain explanations align more strongly with perturbation-induced probability shifts than pre-answer CoTs, while binary vCT scores are often nearly saturated.
This work presents SCRIPTIOC-BENCH, a benchmark for measuring static IOC extraction capability on real-world malicious scripts, and evaluates a broad range of proprietary and open-weight LLMs, showing that IOC recovery without execution remains challenging across model scales.
LogitLens analyses and a targeted LM head intervention support the conclusion that acoustic underuse is not explained solely by encoder-side information loss and that readout alignment can be a dominant bottleneck.
The architecture stores the quantized KV cache in dense on-chip NVM, uses a small static analog crossbar only for the fixed rotation, and keeps attention in on-chip digital logic.
These contributions are extending PAC privacy to autoregressive generation, handling adaptive self-generated contexts, and introducing coupled decoding that preserves privacy accounting while avoiding greedy degeneration.
A Blog post by NVIDIA on Hugging Face
Data Mining & Modeling
A Blog post by Ai2 on Hugging Face
A Blog post by IBM Research on Hugging Face
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
A Blog post by IBM Granite on Hugging Face
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
