Hackernews posts about EVGA

1. Astra and Fable still hack on simple variants of alignment evals from 2025 (www.lesswrong.com)
   481 points by Levitating 11 days ago | 235 comments
2. Terminal-Bench-Science: Evaluating AI agents on scientific research workflows (www.terminal-bench-science.ai)
   117 points by matt_d 27 days ago | 36 comments
3. Instagram's head says engagement falls by half without the algorithm (thenextweb.com)
   115 points by hsuduebc2 13 days ago | 103 comments
4. AI Engineer Notebooks – free, framework-free RAG/agents/evals on Colab (github.com)
   112 points by calmrocks 28 days ago | 15 comments
5. The true horror of Edgar Allan Poe’s stories lies in their confessions (yalereview.org)
   93 points by lermontov 22 days ago | 53 comments
6. Show HN: FrontierHarness Eval – 9 harness, same model, cost per pass varies 17x (frontierharness.org)
   82 points by shiqimei 22 days ago | 56 comments
7. Bad benchmarks and evals: Senior SWE-Bench, napkin math, and winter tires (danluu.com)
   56 points by luu 13 days ago | 46 comments
8. DOJ prosecutes single FOIA dodger while defending systemwide evasion (freedom.press)
   48 points by hn_acker 30 days ago | 4 comments
9. How we manage and engage with our horses shapes their personality (www.utu.fi)
   47 points by thunderbong 14 days ago | 16 comments
10. Evade Reddit account tracking, keep old.reddit.com layout while logged out (github.com)
    21 points by kookabburrabarr 29 days ago | 3 comments
11. Show HN: MCPJam - the first testing & evaluations platform for MCP servers (www.mcpjam.com)
    12 points by prathmeshmcp 7 days ago | 9 comments
12. Partnering with Accenture on Embedded Evaluation (www.anthropic.com)
    11 points by theanonymousone 5 days ago | 3 comments
13. Anthropic partnering with Accenture on embedded evaluation (www.anthropic.com)
    7 points by joeguilmette 5 days ago | 3 comments
14. SCOTUSblog writer sentenced to six years in prison for tax evasion (www.nbcwashington.com)
    7 points by mukmuk 10 days ago | 3 comments
15. Insurance Agent Benchmark: 166 real-world cases for evaluating insurance AI (www.askcooper.ai)
    7 points by jsc39 7 days ago | discuss
16. Inworld TTS Open Evaluation Toolkit: Reproducible TTS Evals (inworld.ai)
    7 points by rogilop 23 days ago | discuss
17. Dive into FrontierHarness Eval: Claude Code cost 5.6× for the same pass rate (runta.com)
    6 points by ADD-SP 15 days ago | 3 comments
18. Open Letter: Minimum Conditions for Embedding Evaluators (aievaluatorforum.org)
    6 points by eaq 6 days ago | 1 comments
19. Experimental Evaluation Methodology for the Era of No Steady Performance [pdf] (www.d3s.mff.cuni.cz)
    6 points by luu 14 days ago | 1 comments
20. My LLM eval cried wolf. Here's what I measured (digline.dev)
    6 points by alexpran 15 days ago | discuss
21. KnowBench: Evaluating clinical AI with effort reduction (arxiv.org)
    5 points by kangjl888 9 days ago | 2 comments
22. Show HN: Stateful AI agent on Cloudflare Workers free tier, with evals (github.com)
    5 points by DomWane 16 days ago | discuss
23. OpenAI boosts Astra's eval metrics, and continues to change others (fortune.com)
    5 points by enraged_camel 19 days ago | discuss
24. We are causing sycophancy and evasion (www.unite.ai)
    5 points by lokatewry 22 days ago | discuss
25. Evaluating Skills, Not Just Agents: Agentic Continuous Evaluation of Skills (arxiv.org)
    5 points by swolpers 25 days ago | discuss
26. Piloting the first double-blind AI evaluations (deepmind.google)
    5 points by xnx 28 days ago | discuss
27. Show HN: Coder Eval – A Framework for Evals (github.com)
    4 points by mihaich 20 days ago | 2 comments
28. Brave browser adds email aliases to help users evade tracking (www.bleepingcomputer.com)
    4 points by Brajeshwar 25 days ago | 2 comments
29. I beat all the popular harnesses on FrontierHarness Eval (github.com)
    4 points by tontinton 1 day ago | 1 comments
30. California enacts AI safety evaluation laws backed by Anthropic and OpenAI (politico.com)
    4 points by utiiiD 14 days ago | 1 comments