Hackernews posts about arXiv.org

ArXiv.ORG is an online repository that provides open access to electronic preprints of completed peer-reviewed papers in physics, mathematics, computer science, and related disciplines.

1. Microsoft Paper: LLMs Corrupt Your Documents When You Delegate (Arxiv.org) (arxiv.org)
   7 points by wuschel 17 days ago | 2 comments
2. Show HN: Apple's SHARP running in the browser via ONNX runtime web (github.com)
   184 points by bring-shrubbery 11 days ago | 46 comments
3. Smart TV tracking raises privacy concerns (2024) (www.ucl.ac.uk)
   51 points by nowflux 22 days ago | 59 comments
4. Show HN: Witchcraft and Pickbrain – fast multi-vector semantic search in Rust (github.com)
   6 points by jacobgorm 28 days ago | discuss
5. Show HN: I ported OmniAID image detection model to Apple's Neural Engine (apps.apple.com)
   4 points by woadwarrior01 17 days ago | discuss
6. Show HN: LLM-Rosetta - Translate LLM API Calls Across OpenAI, Anthropic, Gemini (github.com)
   2 points by Oaklight 24 days ago | discuss
7. Show HN: Springdrift – A persistent runtime for long-lived LLM agents (github.com)
   2 points by s_brady 28 days ago | discuss
8. Show HN: Self-hosted Spanner-like database that speaks Redis (github.com)
   1 points by withinboredom 1 day ago | discuss
9. Ask HN: At ~165k tokens, does Opus 4.6 1M outperform Opus 4.6 200k?
   1 points by consumer451 29 days ago | discuss
10. LLMs corrupt your documents when you delegate (arxiv.org)
    479 points by rbanffy 5 days ago | 201 comments
11. There Will Be a Scientific Theory of Deep Learning (arxiv.org)
    367 points by jamie-simon 20 days ago | 168 comments
12. AI Self-preferencing in Algorithmic Hiring: Empirical Evidence and Insights (arxiv.org)
    334 points by laurex 12 days ago | 178 comments
13. Deterministic Fully-Static Whole-Binary Translation Without Heuristics (arxiv.org)
    289 points by matt_d 1 day ago | 65 comments
14. GLM-5V-Turbo: Toward a Native Foundation Model for Multimodal Agents (arxiv.org)
    163 points by gmays 9 days ago | 32 comments
15. Embedded Rust or C firmware? Lessons from an industrial microcontroller use case (arxiv.org)
    162 points by mrtz 13 days ago | 165 comments
16. ProgramBench: Can language models rebuild programs from scratch? (arxiv.org)
    150 points by jonbaer 7 days ago | 80 comments
17. Refusal in Language Models Is Mediated by a Single Direction (arxiv.org)
    118 points by fagnerbrack 12 days ago | 45 comments
18. Different language models learn similar number representations (arxiv.org)
    115 points by Anon84 20 days ago | 47 comments
19. LLMorphism: When humans come to see themselves as language models (arxiv.org)
    82 points by okey 4 days ago | 58 comments
20. ML supports existence of unrecognized transient astronomical phenomena (arxiv.org)
    73 points by solarist 20 days ago | 59 comments
21. Beyond Semantic Similarity (arxiv.org)
    64 points by 44za12 1 day ago | 15 comments
22. Transformers Are Inherently Succinct (2025) (arxiv.org)
    62 points by bearseascape 9 days ago | 9 comments
23. KV Cache Compression 900000x Beyond TurboQuant and Per-Vector Shannon Limit (arxiv.org)
    46 points by EGreg 23 days ago | 42 comments
24. Prefill-as-a-Service:KVCache of Next-Generation Models Could Go Cross-Datacenter (arxiv.org)
    44 points by matt_d 25 days ago | 1 comments
25. FairyFuse: Multiplication-Free LLM Inference on CPUs via Fused Ternary Kernels (arxiv.org)
    24 points by PaulHoule 1 day ago | 1 comments
26. Hallucinations Undermine Trust; Metacognition Is a Way Forward (arxiv.org)
    19 points by gmays 6 days ago | 7 comments
27. Inexpressibility in Exp-Minus-Log (EML) (arxiv.org)
    17 points by unprovable 9 days ago | 1 comments
28. Superintelligent Retrieval Agent: The Next Frontier of Information Retrieval (arxiv.org)
    16 points by CharlesW 4 days ago | discuss
29. Agentic Harness Engineering (arxiv.org)
    15 points by Anon84 13 days ago | discuss
30. Hallucination Is Inevitable: An Innate Limitation of Large Language Models (2025) (arxiv.org)
    14 points by drob518 10 days ago | 11 comments