AI Research news and analysis on Web Pulse: 8399 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
Researchers developed a power-calibrated statistical framework for logit-based LLM watermarking that establishes explicit relationships between hyperparameters, detection power, and distortion, transforming watermark des…
Researchers propose retrieval-augmented generation and constrained decoding to improve LLM-generated web API invocation code. RAG reduces hallucinations but can introduce unnecessary parameters, while constrained decodin…
Researchers show that incorrect L0 hyperparameter settings in Sparse Autoencoders (SAEs) cause feature mixing and degenerate solutions, preventing proper disentanglement of LLM internal features. They propose a proxy met…
Researchers proposed a method to estimate observation parameters and noise in inverse problems using a diffusion-based prior model within a Bayesian framework. The approach enables optimal estimation of both parameters a…
Researchers introduced Algorithmic Core Extraction (ACE) to isolate compact subspaces, or algorithmic cores, that are necessary and sufficient for a task and recur across independently trained transformers. Across synthe…
Researchers introduced a Gibbs posterior sampler for solving inverse problems using a diffusion-based prior model. The method addresses ill-posed linear inverse problems with Bayesian regularization, offering simplicity …
A new AI-driven vulnerability discovery platform, 0day Rubbish, uses a multi-LLM ensemble to find and disclose critical zero-days. Its first case study reveals a CVSS 9.8 unauthenticated RCE chain in Cisco CUCM 14.0, spa…
A developer explains the technical journey from pressing Enter to receiving a response from a large language model (LLM) like ChatGPT, Gemini, or Claude. The process involves tokenization, embedding, and transformer oper…
Researchers introduced agent data injection attacks (ADI), a new category of indirect prompt injection that disguises malicious data as trusted data to trick AI agents into executing unintended actions. The attacks bypas…
Meta Platforms shares fell 11% in June 2026 after the company raised its 2026 capital expenditure forecast to $125-145 billion for AI infrastructure, spooking investors despite strong revenue growth. The tech giant also …
A developer argues that the real bottleneck in AI development is not GPU shortages but a fundamental issue identified in 2007: the lack of efficient algorithms for training large neural networks. The post highlights that…
Kazakhstan unveiled its national pavilion at the AI for Good Global Summit 2026 in Geneva, showcasing its digital ecosystem under the slogan 'Kazakhstan - AI and Digital Hub of Eurasia.' The pavilion highlights the count…
Ant Group's embodied-AI unit Robbyant open-sourced LingBot-Vision, a family of self-supervised Vision Transformers designed for dense spatial perception, on Hugging Face under Apache-2.0. The 1.1B-parameter flagship mode…
A developer tested four Chinese AI models—DeepSeek, Qwen, Kimi, and GLM—across coding, reasoning, creative writing, and Mandarin tasks. DeepSeek V4 Flash emerged as a top pick for its balance of performance and cost at $…
ForkMind, a new open-source tool that treats LLM context windows like Git repositories, launched to help developers branch, debug, and compare AI conversations locally. The tool captures every LLM call into a local direc…
Researchers at Nagoya University led by Masashi Kato and Takumi Kagawa found that luteolin, an antioxidant in vegetables like celery and broccoli, prevented hair graying in mice. The study showed luteolin preserved endot…
Polysemanticity in neural networks arises from superposition, where a single neuron activates for multiple distinct inputs due to insufficient neurons. In language models, this enables efficient representation of compoun…
Google DeepMind's mechanistic interpretability team proposed a pragmatic framework for validating interpretability tools using proxy tasks, demonstrating its effectiveness by subtracting an "eval-awareness" vector from C…
Runway opened its first office in Paris with a 10-person team and a $30 million investment, positioning the hub as a research center for world models and physical AI. The move reflects the company's strategy to recruit t…
Chinese AI startup DeepSeek has begun developing its own AI chips focused on inference workloads to reduce reliance on NVIDIA and cut costs, sources told Reuters. The project is in early stages and has been underway for …