LIVE · refreshes every 20 min
updated Nov 14, 07:23 UTC
110010.art(ificial intelligence)
AI products, research, and launches — clustered and ranked, not just re-blogged.
110010.art / archive

Research

Oct 21
24d ago A Hybrid TFT-NBEATS framework for robust multi-horizon photovoltaic power forecasting arXiv cs.LG → story
24d ago A multi-granularity interaction network for relational triple extraction arXiv cs.CL → story
24d ago Activation steering improves ASR performance and demographic fairness via supervised optimization; paired activation subtraction shows no benefit arXiv cs.CL → story
Oct 20
24d ago Autonomous AI advances human knowledge frontier by conducting problem-driven scientific discovery; Sci paper outlines autonomous research approach arXiv cs.AI → story
25d ago A conditional no-hair theorem for transformers shows that a transformer’s memory of its prompt decays as a power law with token position. arXiv cs.AI → story
25d ago Adversarial Intent Resolution resolves ambiguity in text-to-SQL by using a multi-agent framework that probes competing interpretations without task-specific training arXiv cs.AI → story
25d ago Agents' last doubt: knowing when to act and when to step back in tool-using agents arXiv cs.AI → story
25d ago Beyond explicit intent: discovering and shaping artistic intent through human-AI co-creation arXiv cs.AI → story
25d ago Cache2LoRA: Understanding and improving context internalization arXiv cs.AI → story
25d ago Compact open-weight LLMs can recover missing mobility events from incomplete air-travel trajectories; explicit prompts improve recovery, in a controlled evaluation using air trips arXiv cs.AI → story
25d ago Controlled LLM workflow explains risk score differences using similar historical transactions and SHAP evidence, with reference-based reasoning and deterministic validation. arXiv cs.AI → story
25d ago Cortex: a fixed-point theory of governed coding agents; formalizes requirement closure on a finite lattice for a sound, monotone validation rule with persistent evidence arXiv cs.AI → story
25d ago Distillation, self-training, and the fragility of learned chemical reasoning arXiv cs.AI → story
25d ago Doing less with more: a first-principles evaluation of when agentic computing is preferable to local computation, REST, SOAP, and MCP arXiv cs.AI → story
25d ago From sec filings to financial signals: evaluating LLMs on stock market reactions and analyst forecasts arXiv cs.AI → story
25d ago GardenBench: a lightweight daily evaluation of LLM capabilities arXiv cs.AI → story
25d ago Joint, decoupled, or agent: which multimodal alignment paradigm wins across datasets and budgets. arXiv cs.AI → story
25d ago LogicVal: a logic-formalized benchmark for constraint satisfaction in text-to-image generation arXiv cs.AI → story
25d ago Mirror: a self-supervised video-to-action model for scalable embodied agent planning arXiv cs.AI → story
25d ago Not every user says it all: training tool-using agents with weak users reduces policy success and does not improve performance. arXiv cs.AI → story
25d ago Persistent memory improves first-attempt success in an LLM agent solving 668 Chinese-character drawing tasks arXiv cs.AI → story
25d ago Social media shows accumulative AI risks through repeated use and recurring decisions, according to a position paper. arXiv cs.AI → story
25d ago SteerSearch introduces mid-response steering to redirect search-augmented language models during ongoing reasoning and retrieval to address answer-target anchoring. arXiv cs.AI → story
25d ago TorchTPU enables native PyTorch execution on Google TPUs with an eager/compiled hybrid architecture for seamless model execution and TPU migration arXiv cs.AI → story
25d ago Training-free prompt selection for efficient reinforcement learning with partial rollouts arXiv cs.AI → story
25d ago Unified multiparadigm fusion intelligence for rare disease diagnosis arXiv cs.AI → story
25d ago Who speaks, who authorizes, who acts? Understanding trust boundaries in multi-agent LLMs arXiv cs.AI → story
25d ago World of SWECraft: agent collaboration at scale in simulated software worlds arXiv cs.AI → story
Oct 19
25d ago Teaching AI about symmetry could improve reliability of prenatal diagnoses, using MRI-based movement tracking to aid earlier detection of developmental anomalies. MIT News (AI) → story
26d ago Safetyedit: Sparse internal targets with prompt-wide control for LLM jailbreak backdoors arXiv cs.CL → story
26d ago SAG2: embedding-free multi-hop retrieval with BM25 and JEV arXiv cs.CL → story
26d ago Sparse autoencoders can induce local geometry to mitigate catastrophic forgetting. arXiv cs.LG → story
26d ago Structure-aware spatial kernels for Gaussian process distributed port-Hamiltonian systems arXiv cs.LG → story
26d ago ToMStories: interpreting theory-of-mind in synthetic narratives arXiv cs.CL → story
26d ago TRACE: Transformer residual activation codebook encoding for communication-efficient distributed edge inference arXiv cs.LG → story
26d ago Trajectory-guided steering mitigates spurious feature bias in clinical classification using prediction trajectories across layers arXiv cs.CL → story
26d ago Trust but Verify: redesigns region-inhomogeneous forward dynamics for discrete diffusion arXiv cs.LG → story
26d ago Uncertainty measures RAG hallucination risk only when retrieved evidence constrains the answer; poor retrieval coverage leads to higher risk of unsupported claims arXiv cs.CL → story
26d ago When is a large language model a good calibration agent? Physics-labeled prompting for budgeted engineering simulator calibration arXiv cs.LG → story
26d ago Who do perspectivist models predict well? Annotator agreement, sparsity, and performance arXiv cs.CL → story

Older items: monthly archive →