Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
A new online converter tool demonstrates how numbers are stored across GPU floating-point formats FP32, FP16, BF16, FP8, and FP4, revealing the precision tradeoffs critical to modern AI workloads. The tool, explained in …
Open-weight AI models trail closed-weight frontier models by approximately four months in general capability, according to the Epoch Capabilities Index, which aggregates scores from multiple AI benchmarks into a single s…
ONNX v1.22.0, published June 30, introduces Opset 27 with LinearAttention and CausalConvWithState operators that enable modern hybrid LLMs to export cleanly to ONNX for the first time, alongside WebAssembly-based model v…
A developer built voku/agent-loop, a governed workflow for coding agents that distinguishes between temporary context, evidence, proposed learning, and approved project guidance. The system requires explicit work briefs …
Whatnot, the livestream shopping platform, acquired AI startup Shaped to integrate real-time recommendation and search technology, aiming to enhance personalization and expand into new product categories. The deal positi…
Amazon is testing software to automate warehouse staffing decisions, but some managers are overriding the system, according to internal documents and Slack conversations reviewed by Business Insider. The tech giant aims …
Wi-Fi signals can be used with AI to detect movement, presence, and activities like walking or falling without cameras. The technology analyzes how radio waves change when people move, offering privacy-preserving sensing…
OriginBlame, a new data provenance system described in an arXiv paper, reduces dataset-level over-deletion from 101x to 1.3x when honoring data-removal requests, enabling model trainers to implement unlearning algorithms…
A new machine learning system developed by Queensland University of Technology researchers has flagged more than 250,000 cancer research papers as potentially fraudulent, according to a study published in The BMJ. The AI…
Tyred, a startup developing an AI platform that predicts bicycle breakdowns before they occur, has raised £2.5 million in funding. The company aims to help cyclists avoid unexpected mechanical failures by using predictiv…
A new data provenance system called ob enables record- and token-level tracing for precise data deletion, reducing over-deletion from 101x to 1.3x in tests on 219,555 Wikipedia pages. The system adds only 1.3-4.0% overhe…
Researchers propose using imitation learning to train reinforcement learning agents for cyber-defense, enabling them to predict attacker actions in partially observable networks. By integrating neurosymbolic methods with…
A new study finds that denoising audio with SAM-Audio preprocessing paradoxically worsens zero-shot automatic speech recognition (ASR) performance. Researchers tested five OpenAI Whisper variants on noisy Bengali and Eng…
A new hybrid AI model integrating Radial Basis Function reconstruction, neural networks, and partial differential equation solvers outperforms traditional methods in generating dense physical fields from sparse data, tes…
A new multi-modal dataset from Catalonia, Spain, includes around a million side-scan sonar tiles, bathymetric maps, and co-registered optical images from autonomous underwater vehicle surveys, with about 36,000 tiles man…
A new diagnostic called 'operator-on-F' reveals that traditional reward-prediction metrics in reinforcement learning are poor indicators of planning performance, with operator error showing a Spearman correlation of -0.9…
A new machine-learning framework using a multi-armed bandit algorithm dynamically selects inductive generalization strategies for the IC3 hardware model checker, solving 26 to 50 more cases than traditional approaches an…
New research shows that incentivizing diversity in AI exploration through a representation-based bonus derived from hidden states of pre-trained models boosts verifier efficiency by over 50% for the Qwen-2.5-14b-Instruct…
A new universal operator approximation theorem for encoder-decoder architectures, developed by researchers, challenges traditional constraints by operating independently of compact sets and accommodating normed and metri…
A new approach to polynomial approximation using one-sided weighted deep polynomials promises precision for asymmetrical functions, with numerical tests on Black-Scholes option-pricing functions showing smaller uniform e…