A live pulse on AI, between issues: what researchers are reading, what builders are shipping, and what everyone is arguing about. Auto-refreshes throughout the day.
Community-upvoted research from Hugging Face Daily Papers.
A gentle, kid-friendly way to understand the big idea behind a trending paper.
The newest submissions in cs.AI, cs.LG, and cs.CL — raw and uncurated.
AI repos created in the last 30 days, ranked by stars.
Latest releases from the infrastructure everyone builds on.
What's trending on the Hugging Face Hub right now.
What the community is training and evaluating on.
Trending Spaces — AI demos you can run in the browser right now.
AI projects shown on Hacker News in the past two weeks.
Top AI stories on Hacker News from the past week.
Fresh posts from OpenAI, Google DeepMind, Google AI, and Hugging Face.
Six questions from this week's real feeds. Did you keep up?
What is the main limitation of standard vision-language models (VLMs) according to the Mage-VL paper?
We present a reproducible pipeline for mapping Common Vulnerabilities and Exposures (CVEs) to MITRE ATT&CK Enterprise techniques from free-text vulnerability descriptions. Rather than relying on the CWE->CAPEC->ATT&CK d…
Recovering an editable design file from a raster image is a common and costly bottleneck in modern design workflows, yet remains challenging since editability depends on recovering multi-modal attributes, such as typogr…
Standard vision-language models (VLMs) suffer from Moravec's paradox: they excel at complex offline visual reasoning but struggle with simple streaming perception tasks and process them inefficiently. We present Mage-VL…
Relevance is a query-dependent estimate of whether a document or excerpt contains useful evidence. Existing retrieval agents use relevance to select top-k content, but document relevance alone cannot localize, compose,…
Learning deployable manipulation policies is bottlenecked by the scarcity of data that is both high-fidelity and scalable. Real-robot teleoperation is accurate but costly to scale; robot-free UMI capture scales readily,…
We introduce PerceptionBench, a benchmark specifically designed to evaluate the atomic visual perception capabilities of Multimodal Large Language Models (MLLMs). Existing benchmarks often fail to isolate perception: ho…
Long-term memory systems store what a user says in an external store and retrieve it when a related query arrives. This interface rests on an assumption so natural that it is rarely stated: a memory that is needed will…
In the age of foundation models, a model is only as good as its prompt. For this reason, prompt engineering has become an essential technique for improving language model performance. Since video models are currently be…
Paper Playground
Pick a trending paper. We'll turn its research summary into a tiny story—without pretending the paper says more than it does.
Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion
We present a reproducible pipeline for mapping Common Vulnerabilities and Exposures (CVEs) to MITRE ATT&CK Enterprise techniques from free-text vulnerability descriptions. Rather than relying on the CWE->CAPEC->ATT&CK d…
Jul 28
On-policy distillation (OPD) grounds token-level supervision in the student's own trajectory, yet suffers from prefix failure: once the student commits to a wrong reasoning direction, all subsequent generation builds on…
Jul 28
Generalist manipulation policies increasingly take the form of action-chunking flow policies built on large pretrained backbones. Such chunks run open-loop, so the policy cannot react to sensory input arriving mid-execu…
Jul 28
Mixture-of-Experts (MoE) variants of Low-Rank Adaptation (LoRA) route every token to a fixed number of experts $k$. Tokens differ in how uncertain the model is about them, so a single k over-spends on easy tokens and un…
Jul 28
Enhancing classification performance in mammography remains a persistent challenge across both small curated datasets and large-scale clinical cohorts. Conventional transfer learning approaches often neglect dataset-spe…
Jul 28
We present VetClaw, an edge-cloud multimodal agentic system for early veterinary disease screening. VetClaw uses a camera module as an edge sensing device and sends captured images, together with optional symptom descri…
Jul 28
Computer-use agents (CUAs) increasingly act through desktop GUIs to complete long-horizon tasks. Current benchmarks primarily measure end-task success or single-frame grounding. Neither isolates whether a model can reco…
Jul 28
Much like humans benefit from guidance while learning, reinforcement learning algorithms may benefit from additional supervision beyond rewards. Leveraging additional information during training to learn better represen…
Jul 28
Technological races create tension between speed and safety: actors may gain by moving faster than competitors, even when risky development is harmful. This is prominent in debates about artificial intelligence (AI), wh…
autonomous red teaming platform; multi-agent offensive-security meta-harness
The open-source AI workbench for scientific research
Think with AI beyond the chat box. A shared canvas for handwriting, equations, diagrams, and spatial reasoning.
An arbitrage bot is a smart contract connected to an external automation script that controls its operation.
An open-source, model-agnostic AI workbench for scientific discovery.
GPU worker client for the Talos network. Pairs with your Talos account, serves open-model inference jobs over a WebSocket, and reports uptime for payouts.
AI SRE AgenticOps for Kubernetes and cloud infrastructure.
runs anywhere. uses anything
langchain-ai/langchain
ggml-org/llama.cpp
ollama/ollama
vllm-project/vllm
huggingface/transformers
Patch release: v5.14.1
pytorch/pytorch
PyTorch 2.13.0 Release
image text to text
image text to text
text generation
image text to text
text generation
text generation
text to image
text generation
Live demo
Live demo
Live demo
Live demo
Live demo
Live demo
Live demo
Live demo
Show HN
Show HN
Show HN
Show HN
Show HN
Show HN
Show HN
Show HN
OpenAI
Hugging Face
Google AI
Hugging Face
Google AI
Google AI
Hugging Face
OpenAI