Search the full wire by company, model, lab, or keyword. Every story we have ever aggregated.
Earlier this week, a picture that seemed to show Kentucky Senator Mitch McConnell covered in tubes in a hospital bed in a state of extreme distress. It turned out to be an AI-generated fake.
Kenton Varda banned AI-generated PR/commit messages after finding them missed high-level context while restating obvious code details.
More young girls sue X over Grok CSAM; X accused of shielding child predators.
Elon Musk's tech company released the newest version of Grok on Wednesday, promising a cheaper, more efficient alternative to other powerful AI models.
General Intuition is betting millions of hours of video game data can train the foundation models for physical AI, making it easier to build smarter robots with minimal real-world data.
The feature can do things like apply cinematic relighting to brighten up a dark clip, swap out a plain background for something fun, or add artistic styles to videos.
Deep learning approach for structure-property reasoning in chemistry/materials that preserves domain-native structural information and scientific constraints.
Co-LMLM extends limited-memory language models with continuous vector keys for flexible knowledge retrieval, improving over discrete relational KBs.
Analysis of transformer linearization via state update design; introduces sink tokens to improve long-context inference while preserving model quality.
Causal trajectory analysis framework for LLM-based agent optimization, extracting root causes from noisy execution traces to improve long-horizon policies.
Jailbreak system uses agentic code generation to bypass database drivers and directly read storage files for analytical workloads.
Institutional red-teaming methodology proves deployment rules causally shape multi-agent AI safety via IABench-CA with 33k+ games across seven model populations.
Timestep weighting and advantage-based replay improve feedback efficiency of diffusion RLHF, reducing human evaluation bottleneck.
Agon trains reasoning models via competitive RL where two agents grade each other's solutions, incentivizing better thinking vs. longer traces.
When it comes to achieving artificial general intelligence (AGI), large language models just don’t have what it takes. Models like ChatGPT and Claude are great at text, but they’re less skilled at understanding how things actually move through space and time — an essential skill for producing intelligence that generalizes. That gap, it turns out, might be filled by gaming data. That’s the bet behind General Intuition, a […]
Lightweight AI framework digitizes paper ECGs and screens for myocardial infarction in remote clinics with limited connectivity.
NOTES combines neural operators, dimensionality reduction, and evolutionary optimization for robust, transferable PDE-constrained inverse design.
Study on how ML models generalize from small to large inputs and sketching techniques for evaluating models on unseen input sizes.
Analysis of RoPE frequency selection in transformers showing data-driven matching between positional frequencies and training data structure.
SkillCenter releases 216,938 structured skills across 24 domains from peer-reviewed sources and GitHub for autonomous agent execution.
AdaPrefix-GRPO adapts solution prefix length during training to maintain 50% success rate, improving gradient signal on hard reasoning tasks.
MedPMC framework extracts high-fidelity multimodal clinical data from PubMed Central for medical foundation model development.
PeTeR applies post-training Wasserstein-robust optimization to probabilistic circuits for improved generalization under distribution shift.
Analysis of Bielik LLMs (1.5B-11B) showing activation dispersion in MLP layers predicts entity familiarity and factual reliability across domains.
DiaLLM studies English dialect generation in open-weight LLMs via continual pretraining on ICE, comparing alignment strategies across Australian, Indian, British English.
Analysis of classifier-free guidance overstabilization in diffusion samplers with proposed terminal-fitted repair for high-guidance regimes.
Survey of 1,250 arXiv papers (2024-2026) categorizing recursive self-improvement in AI systems by loop closure and improvement target.
Meta is adding a new safeguard to stop people from secretly recording others with its AI glasses. But the update comes as the company continues to expand how much personal data its AI products collect and use.
RL post-training composes primitive skills into novel reasoning strategies beyond base model capabilities in controlled symbolic environments.