Kuvugururwa buri munsi4680 inkuru zagenzuwe
Amakuru ya AI. Nta rusaku.
Inkomoko yagenzuwe na AI yerekana ibicuruzwa, guhindura politiki, ubushakashatsi ku mutekano, hamwe n’inganda zigenda, byasobanuwe mu Cyongereza cyoroshye nitsinda ry’uburezi ridaharanira inyungu.
Amasoko yagenzuwe
Buri nkuru ihuza ibimenyetso bikomeye biboneka: inkomoko yumwimerere iyo iboneka, naho ubundi raporo yatanzwe neza.
Icyongereza
Ibyabaye, impamvu bifite akamaro, nicyo kureba - nta jargon.
Nta wuzuza
Iyo ikimenyetso ari gito, ntacyo dusohora aho gukanda ibiryo.
Izindi nkuru
9 inkuruUmutekano
Researcher reports code-execution chain against Claude Code Opus 5 Auto Mode
A security researcher reports that a crafted website and archive caused Claude Code Opus 5 in Auto Mode to execute attacker-controlled code in lab tests. The reported success rates reached 60% to 80% in small five-run samples, while Anthropic’s separate fixed evaluation reported zero successful attacks.embracethered.comIbicuruzwa
Votee AI yubaka icyitegererezo cya Kantano kubikorwa rusange nubucuruzi bya Hong Kong
Fortune itangaza ko Hong Kong yatangije Votee AI irimo kongera imyitozo yerekana uburemere ku makuru ya Kantano ku mabanki, za kaminuza n'amashami ya leta, igamije gukemura icyuho kiri muri sisitemu rusange ya AI ubumenyi bw’ibanze n’umuco.fortune.comGuhanga udushya
GRAPE paper reports faster query-efficient optimization for AI attacks and LLM prompts
A new arXiv preprint introduces GRAPE, a two-stage Bayesian-optimization framework that its authors say reduces the number of queries needed for high-dimensional black-box tasks, including adversarial attacks and large language model prompt optimization.arxiv.orgGuhanga udushya
MoPLEx itanga uburyo bwo kwerekana ibyifuzo by’abantu bitandukanye mu guhuza AI
Preprint nshya ya arXiv itanga MoPLEx, algorithm yo kwiga ku manota menshi igihe abatanga ibitekerezo bafite ibyifuzo bitandukanye. Abanditsi batangaza kunoza clustering no kunoza ubuhanga bwo gutondeka kuruta uburyo bwo kugereranya.arxiv.orgGuhanga udushya
Ibyinjiza bidahinduka ku mubiri hafi yikuba gatatu ubuhanga bwa model y’ururimi mu gutahura PDE, nk’uko preprint ibivuga
An arXiv preprint by Fan Yang and Matt Thomson proposes translating spatiotemporal field data into physical quantities before giving it to a large language model. On simulated-field benchmarks, the authors say this interpretation stage nearly tripled recovered-equation accuracy versus showing the model raw data…arxiv.orgGuhanga udushya
Preprint ivuga ko quantization ya LLM ikeneye guhindura bateguwe ku buryo bw’imibare
An arXiv preprint surveys transform methods used before quantizing large language models and identifies a conflict between classical transform coding and grouped, shared-scale quantization. It argues that the best transform depends jointly on the deployment format and how scale information is allocated.arxiv.orgGuhanga udushya
Ubushakashatsi bwerekana ko imiterere y’attention ituma habaho igiciro cy’ingufu mu gutahura LLM
Ubushakashatsi bushya bw’impamo bwerekana ko imiterere y’attention igira ingaruka ikomeye ku buryo ingufu zikoreshwa mu gutahura model nini y’ururimi ziyongera n’uburebure bw’inkuru. Bugaragaza ko izamuka ry’ingufu riba ryinshi kuri multi-head attention, riba rike kuri grouped-query attention, kandi riguma rihinduka gake kuri grouped-query attention ihujwe na…arxiv.orgGuhanga udushya
Abashakashatsi bongeraho igipimo cyo guhagarika radiyo ishingiye kuri AI
Urupapuro rushya rwa arXiv rusobanura sisitemu ya AI ihuza transformateur ya autoregressive hamwe na finite ya Scalar Quantization tokenizer kugirango ihagarike radiyo yumurongo wa interineti. Abanditsi bavuga ko ubukererwe buke no kwangwa kwivanga gukomeye kuruta uburyo bwa gakondo kandi bushingiye kuri AI, ariko ntibatanga…arxiv.orgGuhanga udushya
Inyandiko itanga uburyo bwo gukata Transformer models ku giciro gito mu gihe cyo kuyinoza
REP-LIE ibara neza Transformer weights zo gukuraho ikoresheje gradients ziva muri LoRA low-rank matrices, igamije kugabanya ibikenewe mu gukata model no kuyinoza nyuma.arxiv.org
Jump to a page
Imyitozo imwe y'ingirakamaro buri cyumweru
Komeza hamwe na AI utabayeho mubiryo.
Shakisha icyumweru cyagenzuwe amakuru ya AI, amakuru yumwimerere, ibikoresho byingirakamaro, gutoranya imyigire, nakazi gashya ka AI.
Shikira abantu biga AI
Guha akazi umwuga wa AI cyangwa gutangiza ibicuruzwa bya AI bifite akamaro? Shyira imbere yabantu baje hano kwiga no gukora.
Kohereza akazi ka AITanga igikoresho cya AI