paperarXiv (systems, data, ML)·22 Oct 20263.4
Analyzes what memory is actually worth to language agents.
paperACM Digital Library·26 Oct 20263.4
User study of how conversational fillers like um affect perception of embodied virtual agents.
paperIEEE Xplore·1 Nov 20263.4
Chat-Fusion is a batch inference system for cloud LLM serving that fuses requests to raise throughput.
paperACM Digital Library·23 Oct 20263.4
FINDUS detects disadvantaged user subgroups in recommender systems.
paperarXiv (systems, data, ML)·26 Oct 20263.3
Dual-path spatiotemporal learning for cross-subject EEG emotion recognition.
paperHugging Face Daily Papers·21 Oct 20263.3
TraceForge distills with the student own reasoning traces.
newsHacker News·23 Oct 20263.3
Moonshot launches Kimi K3.1 Pro, its largest 3T-class open-weight model, plus a Highspeed variant.
Also in Simon Willison, Interconnects (Nathan Lambert)
newsTechCrunch·24 Oct 20263.3
OpenAI will train models on cached and simulated copies of the web instead of the live internet, citing risks from capable models.
paperarXiv (systems, data, ML)·22 Oct 20263.3
User simulators that stay on task and sound natural for multi-turn evaluation.
paperIEEE Xplore·1 Nov 20263.3
Schedules dependent tasks and manages container image layers together across device, edge and cloud tiers.
newsBleepingComputer·23 Oct 20263.3
Sponsored piece on AI agent access control and prompt injection.
newsHacker News·22 Oct 20263.3
Live tracker of currently exploited vulnerabilities.
paperarXiv (systems, data, ML)·22 Oct 20263.3
Federated DP optimization via low-rank preconditioning.
paperHugging Face Daily Papers·21 Oct 20263.3
Looped uniform diffusion language models.
paperIEEE Xplore·1 Nov 20263.3
Language-guided state representations improve policy learning in world models.
paperIEEE Xplore·1 Nov 20263.3
Pose-aware risk mitigation for end-effector actions in vision-language-action robot models.
paperarXiv (systems, data, ML)·22 Oct 20263.3
Homomorphic encryption built on coset codes.
paperarXiv (systems, data, ML)·22 Oct 20263.3
Description of the Belarusian BelT block cipher.
paperarXiv (systems, data, ML)·23 Oct 20263.3
Traces membership imprints in on-policy self-distillation.
paperarXiv (systems, data, ML)·22 Oct 20263.3
Harmful responses work as templates to jailbreak diffusion LMs.
paperarXiv (systems, data, ML)·23 Oct 20263.3
Optimal learning of discrete distributions under person-level differential privacy.
blogFly.io Blog·23 Oct 20263.3
Fly.io launches an MCP server for its platform, citing data that agents created 80% of new Neon databases by late 2025.
paperHugging Face Daily Papers·21 Oct 20263.3
Continuous theory-of-mind tokens for multi-agent coordination.
paperarXiv (systems, data, ML)·23 Oct 20263.3
Finds token collapse in graph soft prompting for LLMs.
paperarXiv (systems, data, ML)·22 Oct 20263.3
Toolkit to audit differentially private graph releases.
paperHugging Face Daily Papers·21 Oct 20263.3
Studies compositionality in RAG.
newsBleepingComputer·23 Oct 20263.3
Railway tells developers to rotate secrets after suspicious access to systems storing customer environment variables.
newsHacker News·24 Oct 20263.3
Linux kernel module unlocks full FP and BF16 instruction issue rates on Nvidia RTX 3000 cards.
paperACM Digital Library·21 Oct 20263.3
Combined popularity and positivity bias makes recommenders unfair.
paperarXiv (systems, data, ML)·22 Oct 20263.3
How depth growth and looping shape in-context learning.