Smart News

Hackernews posts about MMLU

RSS feed for MMLU
  1. 01
    The Two MMLU Scores: What a Benchmark Name Does Not Fixzatona.dev
    ▲ 8gmays21d💬 3
  2. 02
    Show HN: Swift-Qwen3.8-27B, -58.3% thinking, x1.95 speed, accuracy of xhighhuggingface.co
    ▲ 33kisjovan19d💬 17
  3. 03
    Show HN: Mluva – Dictation and editable rewrites for Omarchygithub.com
    ▲ 31vecera5ddiscuss
  4. 04
    CHERIoT Provides Strong and Usable Isolation Without an MMUqueue.acm.org
    ▲ 2birdculture21ddiscuss
  5. 05
    CHERIoT provides strong and usable isolation without an MMUqueue.acm.org
    ▲ 2fanf224ddiscuss
  6. 06
    How CHERIoT Provides Strong and Usable Isolation Without an MMUqueue.acm.org
    ▲ 1signa1125ddiscuss
  7. 07
    Show HN: Pyxel – A Python retro game engine with built-in art and sound editorsgithub.com
    ▲ 102kitao3d💬 9
  8. 08
    So you want to use OpenRouter?mmoustafa.com
    ▲ 766player8526d💬 207
  9. 09
    Advice to a Beginning Graduate Student (2001)www.cs.cmu.edu
    ▲ 58nicoraga9d💬 21
  10. 10
    So you want to use OpenRouter?mmoustafa.com
    ▲ 8mmoustafa27ddiscuss
  11. 11
    The first mRNA flu shot is now available. Here's who can get itwww.nbcnews.com
    ▲ 4elo200013d💬 2
  12. 12
    How I Extend Claude Codemalucelli.net
    ▲ 4malucelli27ddiscuss
  13. 13
    How to make your agent do its own supportmmoustafa.com
    ▲ 2mmoustafa29ddiscuss
  14. 14
    When Claude CLI asks for feedback, responding authorizes conversation capture
    ▲ 25mmaunder14d💬 3
  15. 15
    Show HN: Crew — Let Claude/Codex/OpenCode agents talk to each othergithub.com
    ▲ 2mmoustafa28d💬 3
  16. 16
    Show HN: Karma Compass – Agentic board management for nonprofitscompass.karmahq.org
    ▲ 2mmurthy5ddiscuss
  17. 17
    MTurk closed this week: how to keep Ground Truth and A2I workflows runningservices.deepen.ai
    ▲ 1mmusa2ddiscuss
  18. 18
    Show HN: Gatling.chat – The Turbo Messenger for LinkedIngatling.chat
    ▲ 1mlukaszczyk6ddiscuss
  19. 19
    Blocklist Whitelist Anything on YouTube or YT Kids
    ▲ 3devanshvarshney12d💬 1
  20. 20
    Clipx – An open-source AI-native creative suite for macOSgithub.com
    ▲ 2imlukeshen4ddiscuss
  21. 21
    Show HN: Run MMLU benchmark on any LLM endpointmmlu.borgcloud.ai
    ▲ 2lostmsu520ddiscuss
  22. 22
    LLM Comparison/Test: 25 SOTA LLMs (Including QwQ) Through 59 MMLU-Pro CS Runshuggingface.co
    ▲ 2ororm669ddiscuss
  23. 23
    Basaltlabs Monolith-1.0 – #1 on Last Exam, AIME, GPQA Diamond, MMLU-Probasaltlabs.org
    ▲ 1Topfi79d💬 2
  24. 24
    Graph-based multi-agents smash long-context benchmarks–89% MMLU-Pro on 8B modelsgithub.com
    ▲ 1FeelTheAGI2230d💬 1
  25. 25
    Show HN: AutoThink – Boosts local LLM performance with adaptive reasoning
    ▲ 397codelion495d💬 68
  26. 26
    Show HN: Cactus Hybrid: We taught Gemma 4 to know when it's wronggithub.com
    ▲ 191HenryNdubuaku75d💬 44
  27. 27
    Show HN: Arch-Router – 1.5B model for LLM routing by preferences, not benchmarks
    ▲ 66adilhafeez461d💬 15
  28. 28
    Launch HN: General Instinct (YC P26) – Frontier models on edge devices
    ▲ 63guanming0717122d💬 16
  29. 29
    Show HN: BenchFlow – run AI benchmarks as an APIgithub.com
    ▲ 24xdotli562d💬 1
  30. 30
    Show HN: QwQ-32B APIs – o1 like reasoning at 1% the cost
    ▲ 17ozgune628d💬 3
  31. 31
    LLM Benchmark: Frontier models now statistically indistinguishable
    ▲ 7js4ever289d💬 4
  32. 32
    Show HN: I built the LLM Comparison Tool I wish existedllm-stats.com
    ▲ 7JonathanChavez639d💬 3
  33. 33
    Show HN: Flint – A 30B model fine-tuned for less repetitionspringboards.ai
    ▲ 6thmsmxwll172d💬 2
  34. 34
    Show HN: Claude Code 2.0 router – preference-aligned routing to multiple LLMsgithub.com
    ▲ 4adilhafeez369d💬 1
  35. 35
    Show HN: Forecaster Arena – Testing LLMs on real events with prediction marketsforecasterarena.com
    ▲ 4setrf297ddiscuss
  36. 36
    Show HN: 1.5B LLM routing model that aligns to preferences, not leaderboardshuggingface.co
    ▲ 4honorable_coder444ddiscuss
  37. 37
    Show HN: MALLM – A Multi-Agent Framework for Task Solvinggithub.com
    ▲ 4jpwahle459ddiscuss
  38. 38
    Diffusion LLM may make most of the AI engineering stack obsolete
    ▲ 3victorpiles99208ddiscuss
  39. 39
    Show HN: Model-literals, model-aliases, and preference-aligned routing for LLMsdocs.archgw.com
    ▲ 2honorable_coder379ddiscuss
  40. 40
    Show HN: LLM Benchmarking Suitegithub.com
    ▲ 2Dhyaneesh557ddiscuss
  41. 41
    Show HN: OptiLLMBench – Test how inference optimization tricks scale up LLMs
    ▲ 2codelion595ddiscuss
  42. 42
    Show HN: AIBenchy – Independent AI Leaderboardaibenchy.com
    ▲ 1XCSme229d💬 1
  43. 43
    Show HN: MarginDash – See which AI customers are profitablemargindash.com
    ▲ 1gdhaliwal23231d💬 1
  44. 44
    Show HN: MarginDash – See which AI customers are profitablemargindash.com
    ▲ 1gdhaliwal23234d💬 1
  45. 45
    Ask HN: Are LLMs getting better, how can you tell?
    ▲ 1ahamilton454613d💬 1
  46. 46
    ARCHE3-7B – Sparse Moe with SmartRouter and Foundation Curriculum Training
    ▲ 1OpenSynapseLabs185ddiscuss
  47. 47
    Show HN: Mafia Arena – LLMs play social deduction games against each othermafia-arena.com
    ▲ 1mohsen1278ddiscuss
  48. 48
    Show HN: Arch-Router – Aligning LLM Routing with Human Preferencesarxiv.org
    ▲ 1honorable_coder428ddiscuss
  49. 49
    Show HN: Atlas: Independent Evals and Benchmarking for Generative AI Modelsapp.layerlens.ai
    ▲ 1Arch223508ddiscuss
  50. 50
    Show HN: Tokenkit – Convert LLMs to new tokenizers (incl byte-level Llama/Gemma)github.com
    ▲ 1bminixhofer530ddiscuss

Keyboard shortcuts

j / k
next / previous story
o or Enter
open the story
c
open the comments
s
save for later
h
history & saved stories
?
show this help