imudojuiwọn ojoojumọ3981 daju itan
AI iroyin. Laisi ariwo.
Agbegbe AI ti a ṣayẹwo orisun ti awọn ifilọlẹ ọja, awọn iyipada eto imulo, iwadii aabo, ati awọn gbigbe ile-iṣẹ, ṣalaye ni ede Gẹẹsi nipasẹ ẹgbẹ ẹkọ ti kii ṣe èrè.
Ifowosowopo ti o ni idaniloju
Gbogbo itan ni asopọ si ẹri ti o lagbara julọ ti o wa: awọn orisun atilẹba nigbati o wa, bibẹkọ ti o han gbangba.
English itele
Kini o ṣẹlẹ, idi ti o ṣe pataki, ati kini lati wo - laisi jargon.
Ko si kikun
Nigbati awọn ifihan agbara jẹ tinrin, a jade ohunkohun kuku ju òwú kikọ sii.
Awọn itan diẹ sii
9 awọn itanAtunse
Ijabọ asọtẹlẹ jiometirika oniduro idari jẹ ki titete ọpọlọ-awoṣe diẹ sii ni ọna meji
Atejade arXiv kan sọ pe ṣiṣatunṣe jiometirika iwoye ti awọn aṣoju awoṣe iran lakoko ikẹkọ jẹ ki awoṣe – titete ọpọlọ diẹ sii ni iṣiro, jijabọ 55% ere ibatan ni asọtẹlẹ-ọna meji. Awọn onkọwe pe ifihan ni ibẹrẹ, awọn orukọ áljẹbrà ko si data nkankikan, ati ere jẹ pipa-ọja.arxiv.orgAtunse
Preprint proposes batched parallel decoding to cut training time for compact image-generation models
A four-author arXiv preprint introduces HB-SJD, a batched speculative Jacobi decoding backend for the sample-generation step of visual on-policy distillation. The authors report shorter rollout and end-to-end training times with LlamaGen and preserved quality, but no numbers are given and it is not peer reviewed.arxiv.orgAtunse
Philosophy paper argues machine learning should borrow its standards of proof from clinical translation
A preprint accepted by Studies in the History and Philosophy of Science treats the comparison between medicine and machine learning as a formal analogy rather than a slogan, and uses it to sketch a process-based account of when an ML system deserves trust. It is conceptual: no experiments, thresholds or checklists.arxiv.orgAtunse
Preprint ṣe awọn irinṣẹ itumọ bi iṣoro wiwọn kan, ṣe idanwo lori GPT-2 ati Qwen
Onkọwe-ọkan kan arXiv preprint daba kikọ patching ibere ise, gradients ati awọn ọja Hessian-fekito bi iṣoro wiwọn laini kan, ati awọn ijabọ ti o waye-jade lori eto iṣakoso nkan isere, Tracr, GPT-2-kekere ati Qwen-2.5-7B.arxiv.orgAtunse
Preprint ṣe imọran Dimegilio biba Alusaima ti nlọsiwaju ti a kọ lati awọn ọlọjẹ ọpọlọ ti o leralera
Atokọ 13-onkọwe arXiv ṣapejuwe Ipo Ilọsiwaju Arun, ọna Bayesian kan ti o gbe eniyan kan si lilọsiwaju Alṣheimer, pẹlu aidaniloju, lati aworan tensor tan kaakiri leralera. Awọn onkọwe sọ pe o lu awọn ọna ilọsiwaju ti o wa tẹlẹ lori ADNI; áljẹbrà ko fun awọn nọmba, awọn ipilẹ tabi awọn alaye ẹgbẹ.arxiv.orgAtunse
Preprint reports physics-informed network gains came from one pairing, broke when stacked
A single-author arXiv preprint asks whether published tricks for training physics-informed neural networks combine. On an unsteady cylinder-wake benchmark, most matched an untreated baseline alone, one pair reached 4.1% average relative error against OpenFOAM, and stacking more degraded results sharply.arxiv.orgAtunse
Iwe sọ pe oluṣowo kan ti o gba ikẹkọ nikan lori awọn ṣiṣan sintetiki ti o ṣamọna awọn ami-ami-shot odo
RecPFN, atunyẹwo ẹlẹgbẹ ni SIGIR 2026, jẹ ikẹkọ tẹlẹ lori awọn ṣiṣan sintetiki nikan, lẹhinna ṣe asọtẹlẹ awọn ohun kan atẹle fun katalogi ti ko mọ lati awọn ilana apẹẹrẹ diẹ, laisi atunṣeto. Awọn onkọwe ṣe ijabọ ipo-ti-ti-aworan awọn abajade odo-shot lori awọn ipilẹ gbangba mẹjọ; awọn áljẹbrà awọn orukọ ko si metiriki, baselines tabi datasets.arxiv.orgAtunse
DeltaML-Bench ri asefodi aṣoju iyipada aṣeyọri lori awọn iṣẹ ṣiṣe iwadi ẹrọ-ẹkọ
Ijabọ ala tuntun arXiv pe wiwa-orisun scaffolding dara si awọn abajade GPT-5 lori awọn ibi ipamọ iwadii ẹrọ aipe, lakoko ti awọn atunto boṣewa ṣe afihan ere sipesifikesonu.arxiv.orgAtunse
Preprint proposes answer-level trust checks for physical vision-language model predictions
A new preprint proposes a model-agnostic method for deciding whether individual vision-language answers about physical quantities are trustworthy. Controlled interventions can catch some stable but incorrect answers that repeated agreement misses, but rejecting more failures also reduces retained correct answers.arxiv.org
Alaye ti o wulo ni ọsẹ kọọkan
Tẹsiwaju pẹlu AI laisi gbigbe ninu ifunni.
Gba awọn iroyin AI ti a ṣayẹwo ti ọsẹ naa, data atilẹba, awọn irinṣẹ to wulo, awọn yiyan ẹkọ, ati awọn iṣẹ AI tuntun.
De ọdọ awọn eniyan ti o nkọ AI
Igbanisise ọjọgbọn AI kan tabi ṣe ifilọlẹ ọja AI ti o wulo? Fi si iwaju awọn eniyan ti o wa nibi lati kọ ẹkọ ati ṣe.
Firanṣẹ iṣẹ AI kanFi ohun elo AI silẹ