Hacker News: blackcat201

LLMs Refuse High-Cost Attacks but Stay Vulnerable to Cheap, Real-World Harm

blackcat201 — Tue, 10 Feb 2026 05:19:33 +0000

Article URL: https://expectedharm.github.io/

Comments URL: https://news.ycombinator.com/item?id=46955705

Points: 3

# Comments: 0

Open-Source Alternative to Claude Cowork Desktop App

blackcat201 — Fri, 30 Jan 2026 00:25:00 +0000

Article URL: https://github.com/kuse-ai/kuse_cowork

Comments URL: https://news.ycombinator.com/item?id=46818968

Points: 1

# Comments: 0

Kimi Linear: An Expressive, Efficient Attention Architecture

blackcat201 — Fri, 31 Oct 2025 00:07:36 +0000

Article URL: https://github.com/MoonshotAI/Kimi-Linear

Comments URL: https://news.ycombinator.com/item?id=45766937

Points: 217

# Comments: 47

Testing Image Generation Models with Explainable Human Evaluation

blackcat201 — Wed, 29 Oct 2025 07:35:52 +0000

Article URL: https://tiger-ai-lab.github.io/ImagenWorld/

Comments URL: https://news.ycombinator.com/item?id=45743765

Points: 2

# Comments: 0

What Chess (Might) Taught Us About Programming with AI

blackcat201 — Tue, 24 Jun 2025 15:34:32 +0000

Article URL: https://theblackcat102.github.io/what-chess-taught-us-about-programming-with-ai/

Comments URL: https://news.ycombinator.com/item?id=44367337

Points: 2

# Comments: 0

New comment by blackcat201 in "Gödel Agent: A self-referential agent framework for recursive self-improvement"

blackcat201 — Sun, 13 Oct 2024 05:54:54 +0000

Shameless plug, for anyone who's interested in "self-improvement" agent check out StreamBench[1] where we benchmark and try out what's essential for improvements in online settings. Basically we find feedback signal is vital and the stronger the signal the more improvement you can get if you were able to feed it back to the agent in terms of weights (LoRA) or in-context examples.

[1] https://arxiv.org/abs/2406.08747

A Study on the Impact of Structured Output on Performance of LLMs

blackcat201 — Wed, 07 Aug 2024 01:15:13 +0000

Article URL: https://arxiv.org/abs/2408.02442

Comments URL: https://news.ycombinator.com/item?id=41177267

Points: 2

# Comments: 0

New comment by blackcat201 in "Structured Outputs in the API"

blackcat201 — Wed, 07 Aug 2024 01:10:35 +0000

Do beware on some reasoning task, our recent work[0] actually found it may cause some performance degradation as well as possible reasoning weakening in JSON. I really hope they fix this in the latest GPT-4o version.

[0] https://arxiv.org/abs/2408.02442

Ambidex robot arms by Naver[video]

blackcat201 — Wed, 03 Apr 2024 03:49:52 +0000

Article URL: https://www.youtube.com/watch?v=3LvZyOY6vVk

Comments URL: https://news.ycombinator.com/item?id=39913438

Points: 1

# Comments: 0

New comment by blackcat201 in "7.4 earthquake in Taiwan, 34km depth"

blackcat201 — Wed, 03 Apr 2024 01:33:36 +0000

The standard operation is to stop and check if any machine was out of calibration. So yes

New comment by blackcat201 in "Why host your own LLM?"

blackcat201 — Tue, 15 Aug 2023 13:39:15 +0000

I own my LLM not because I need it now but having the luxury to fall back if openai ran out of money

New comment by blackcat201 in "Do we really need a specialized vector database?"

blackcat201 — Sat, 12 Aug 2023 11:19:40 +0000

I have been following the vector database trend back in 2020 and I ended up with the conclusion: vector search features are a nice to have features which adds more value on existing database (postgres) or text search services (elasticsearch) than using an entirely new framework full of hidden bugs. You could get way higher speedup when you are using the right embedding models and encoding way than just using the vector database with the best underlying optimization. And the bonus side is that you are using a stack which was battle tested (postgres, elasticsearch) vs new kids (pinecone, milvus ... )

Why robotics companies fails (2021)[pdf]

blackcat201 — Tue, 01 Aug 2023 01:26:40 +0000

Article URL: https://www.freshconsulting.com/wp-content/uploads/2021/03/Why-Robotics-Fail.pdf

Comments URL: https://news.ycombinator.com/item?id=36951200

Points: 4

# Comments: 0

New comment by blackcat201 in "Ask HN: Could you share your personal blog here?"

blackcat201 — Wed, 05 Jul 2023 01:23:06 +0000

https://theblackcat102.github.io/

Recently I am ranting the AI trends and some short writeup of things I read

New comment by blackcat201 in "Show HN: Cerelyze: Implement research papers into code 100x faster(Desktop only)"

blackcat201 — Sun, 02 Jul 2023 01:03:51 +0000

This looks pretty interesting! But the landing page has only one sentence : Understand and implement research papers faster, followed by a get in touch button. Care to extrapolate more? The blog button doesn't work as well?

Unpopular Opinions on AI

blackcat201 — Fri, 23 Jun 2023 05:17:15 +0000

Article URL: https://theblackcat102.github.io/short-writeup-on-unpopular-opinions-on-ai/

Comments URL: https://news.ycombinator.com/item?id=36443081

Points: 2

# Comments: 0

Chinese Traditional 3B Chat Model

blackcat201 — Thu, 22 Jun 2023 02:01:19 +0000

Article URL: https://huggingface.co/ikala/bloom-zh-3b-chat

Comments URL: https://news.ycombinator.com/item?id=36427166

Points: 1

# Comments: 0

Short writeup on challenges of LLMs in May 2023

blackcat201 — Sat, 27 May 2023 03:36:53 +0000

Article URL: https://theblackcat102.github.io/short-writeup-on-challenges-of-llms/

Comments URL: https://news.ycombinator.com/item?id=36091867

Points: 3

# Comments: 0

New comment by blackcat201 in "The company behind Stable Diffusion appears to be at risk of going under"

blackcat201 — Sat, 08 Apr 2023 01:45:46 +0000

Note that stability have been funding freelance researcher by providing compute resource such as RWKV[1], Open Assistant, some works by LAION[2] and lucidrains[3]

[1] https://github.com/BlinkDL/RWKV-LM

[2] https://huggingface.co/laion/CLIP-ViT-L-14-laion2B-s32B-b82K

[3] https://github.com/lucidrains/gigagan-pytorch#appreciation

New comment by blackcat201 in "OpenAI is now everything it promised not to be: closed-source and for-profit"

blackcat201 — Wed, 01 Mar 2023 12:13:14 +0000

On the other hand, meta AI research should be renamed as OpenAI. They are the only few big institute who open almost every models they train (galactica, OPT, M2M, wav2vec ...)