15 6 24

Emin Temiz PRO

etemiz

https://pickabrain.ai

AI & ML interests

Alignment

Recent Activity

commented on their article 5 days ago

The AHA Indicator

posted an update 20 days ago

This is a very human aligned fine tune and would score 56 on AHA leaderboard: https://huggingface.co/CWClabs/CWC-Mistral-Nemo-12B-V2-q4_k_m We are going to be using it as one of the ground truths for AHA Leaderboard 2.0 (the next version). We will be able to generate some RL datasets for folks to align their own LLMs with humanity. We will generate answers from best models and worst models and do mixture of agents that combines the answer, and publish results as dataset(s). Things looking bright!

liked a model 25 days ago

mradermacher/Ostrich-32B-Qwen3-251003-i1-GGUF

View all activity

Organizations

None yet

commented on The AHA Indicator 5 days ago

Update to the indicator

https://huggingface.co/posts/etemiz/490866222392672

posted an update 20 days ago

Post

277

This is a very human aligned fine tune and would score 56 on AHA leaderboard:

CWClabs/CWC-Mistral-Nemo-12B-V2-q4_k_m

We are going to be using it as one of the ground truths for AHA Leaderboard 2.0 (the next version).

We will be able to generate some RL datasets for folks to align their own LLMs with humanity. We will generate answers from best models and worst models and do mixture of agents that combines the answer, and publish results as dataset(s). Things looking bright!

liked a model 25 days ago

mradermacher/Ostrich-32B-Qwen3-251003-i1-GGUF

33B • Updated 25 days ago • 4.59k • 1

New activity in mradermacher/Ostrich-32B-Qwen3-251003-GGUF 25 days ago

i1?

#1 opened 25 days ago by

etemiz

posted an update 25 days ago

Post

289

Benchmarked 2 more models today. Any other model you want me to do?

Working on a broader version of the AHA leaderboard. Follow for more quackery :)

replied to their post 26 days ago

This fine tuning would score 56 and be placed 1st in the leaderboard but I didn't add it, I only include full trainings in the leaderboard or (further tunings by the same company):

https://huggingface.co/CWClabs/CWC-Mistral-Nemo-12B-V2-q4_k_m

posted an update 26 days ago

Post

258

O oh! things looking uglier by day

LLM builders in general are not doing a great job of making human aligned models.

I don't want to say this is a proxy for p(doom)... But it could be if we are not careful.

Most probable cause is reckless training LLMs using outputs of other LLMs, and don't caring about curation of datasets and not asking 'what is beneficial for humans?'...

1 reply

updated a model 27 days ago

etemiz/Mistral-Nemo-12B-CWC-Enoch-251014-GGUF

12B • Updated 27 days ago • 737 • 1

liked 2 models about 1 month ago

Arki05/Grok-1-GGUF

316B • Updated Apr 11, 2024 • 12.2k • 67

bartowski/Yi-1.5-34B-Chat-GGUF

Text Generation • 34B • Updated May 12, 2024 • 1.25k • 25

published a model about 1 month ago

etemiz/Mistral-Nemo-12B-CWC-Enoch-251014-GGUF

12B • Updated 27 days ago • 737 • 1

replied to their post about 1 month ago

https://huggingface.co/huihui-ai/Huihui-GLM-4.5-Air-abliterated-GGUF/tree/main/Q3_K_M-GGUF

posted an update about 1 month ago

Post

1056

Another abliteration by huihui which had big positive impact!

huihui-ai/Huihui-GLM-4.5-Air-abliterated-GGUF

@huihui-ai

3 replies

updated 2 models about 1 month ago

etemiz/Ostrich-32B-Qwen3-251003

33B • Updated Oct 9 • 9 • 1

etemiz/Ostrich-32B-AHA-Qwen3-250830

33B • Updated Oct 9 • 3 • 1

replied to their post about 1 month ago

Our leaderboard can be used for human alignment in an RL setting. Ask the same question to top models and worst models and the answer from top models can get +1 score, bad models can get -1. Ask many times with higher temperature to generate more answers. What do you think?

posted an update about 1 month ago

Post

198

Grok 2 is worse than 1 but better than 3. This was already measured using API but now we measured the LLM and the results are similar.

GLM is ranking higher and higher compared to previous versions. Nice trend!