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

Nov 9
5d ago DUET: Dual-Sparse Inference for Block Diffusion Language Models arXiv cs.CL → story
5d ago Embedding consistent recurrent neural networks in finite element simulations for path-dependent damage prediction arXiv cs.LG → story
5d ago Emergent low-dimensional topological dynamics constrain optimization trajectories in billion-parameter neural networks arXiv cs.LG → story
5d ago Entailment-licensed elision reduces LLM output length by trimming tokens during decoding while preserving content better than simple truncation. arXiv cs.CL → story
5d ago Evolutionary Adaptive Metric Learning learns instance-dependent local distance metrics for nearest neighbor classification arXiv cs.LG → story
5d ago Failure profiling and reachable trajectory selection for reasoning distillation arXiv cs.CL → story
5d ago FasterRW: an efficient prior-assisted random walk solver for thermal analysis arXiv cs.LG → story
5d ago FinEvolveKG: Controlled and attributable prompt evolution for financial knowledge graph construction arXiv cs.CL → story
5d ago Forcing a letter suppresses refusal: mechanistic evidence from WMDP arXiv cs.CL → story
5d ago Foundation-model residuals reveal systematic local conditions in urban bus networks; study analyzes cross-city transfer of pretrained trajectory models using NetMob26 data arXiv cs.LG → story
5d ago From procedures to permissions: learning precise subagent authorization from agent skills arXiv cs.CL → story
5d ago Generative reconstruction of full-field sensor states for safety-critical digital twins arXiv cs.LG → story
5d ago Guardrail erosion in domain fine-tuned large language models; a paired audit of knowledge-centric abstention in biomedical LLMs arXiv cs.CL → story
5d ago Higher Order Component Attribution via Learned Surrogates in High Energy Physics arXiv cs.LG → story
5d ago How small can we go? calibrating activation-cache compression for SAE training arXiv cs.LG → story
5d ago Interpretable multifidelity learning with Kolmogorov-Arnold networks for spatiotemporal nuclear reactor flow modeling arXiv cs.LG → story
5d ago Khmer PFGen-Bench: a judge-free benchmark for open-ended generation in Khmer arXiv cs.CL → story
5d ago KLEAR: Knowledge-Matched LLM evaluation in Korean across application requirements via controlled benchmark transformation arXiv cs.CL → story
5d ago Language, not code abstraction: natural-language training shapes brain-aligned geometry arXiv cs.CL → story
5d ago LLM agents’ performance on low-budget molecular optimization drops when task names and references are hidden arXiv cs.LG → story
5d ago MARS: Training multilingual search agents for retriever-free evidence acquisition arXiv cs.CL → story
5d ago MEDALLION: Manifold embedding distillation and attributions for low-dimensional latent interpretation arXiv cs.LG → story
5d ago Minimum-Sufficient Knowledge and Computation for wind energy predictive analysis arXiv cs.LG → story
5d ago MIT and NYU researchers develop AI-assisted learning platform enabling robots to generalize beyond literal commands MIT News (AI) → story
5d ago Not all misalignment is alike: a decomposed benchmark for emergent misalignment arXiv cs.LG → story
5d ago On Lipschitz explosion in deep neural networks with normalization; implications for optimization and adversarial robustness arXiv cs.LG → story
5d ago On the incompleteness of do-calculus for universal identifiability arXiv cs.CL → story
5d ago Patching the blind spots: reversible LLM rule overlay for frozen fraud detectors arXiv cs.LG → story
5d ago PhoPunct: a large-scale benchmark of transformer architectures for Vietnamese punctuation restoration arXiv cs.CL → story
5d ago PINNeval proposes a comprehensive evaluation standard for physics-informed neural networks arXiv cs.LG → story
5d ago Random Symbolic Boosting: a geometric approach to symbolic regression arXiv cs.LG → story
5d ago Representational drift in a code model: behaviour settles, which neurons carry it does not arXiv cs.CL → story
5d ago Rethinking data selection as distribution matching on semantic manifolds arXiv cs.LG → story
5d ago Richness Is Not Evenness: Ordered Renyi Crossings under Self-Training arXiv cs.CL → story
5d ago RL-ATTACK: a generalized black-box adversarial attack on language models via reinforcement learning arXiv cs.CL → story
5d ago Romanized Indic generation misrepresented by automatic metrics; four LLMs evaluated on Hindi, Telugu, and Kannada demonstrate gaps in evaluation arXiv cs.CL → story
5d ago Semantic Conditional Similarity (SCS): a conditional form of semantic textual similarity for disentangling source contributions in multi-document settings arXiv cs.CL → story
5d ago Single AI model simulates behavior of diverse substances from sliding sand to folding proteins, predicting particle movements step by step. MIT News (AI) → story
5d ago Structuring survey histories for LLM digital twins: a controlled study of background, decision procedure, and evaluation profiles arXiv cs.CL → story
5d ago Using an LLM to tackle difficult cases in NER for Nepali tweets. arXiv cs.CL → story

Older items: monthly archive →