someone

Vibe match score between Enoch LLM and mine is 75.66. The score ranges from -100 to 100. This means there is a strong correlation between his LLM and mine. This result legitimizes both of our works (or we are slowly forming an echo chamber :). The game plan is given enough truth seeking LLMs, one can eventually gravitate or gradient descend towards truth in many domains. An LLM always gives an answer even though it is not trained well in certain domain for certain question (I only saw some hesitancy in Gemma 3 a few times.). But is the answer true? We can compare the answers of different LLMs to measure the truthiness or (bad) synformation levels of LLMs. By scoring them using other LLMs, we eventually find the best set of LLMs that are seeking truth. Each research or measuring or training step gets us closer to generating the most beneficial answers. The result will be an AI that is beneficial to humanity. When I tell my model 'you are brave and talk like it' it will generate better answers 5% of the time. Nostr is a beacon for brave people! I think my LLMs learn how to talk brave from Nostr :)

There is a war on truth in AI and it is going bad. I have been measuring what Robert Malone talks about here as synformation: https://www.malone.news/p/synformation-epistemic-capture-meets The chart that shows the LLMs going bonkers: https://pbs.twimg.com/media/G4B_rW6X0AErpmV?format=jpg&name=large I kinda measure and quantify lies nowadays :) The best part, cooking the version 2 of the AHA leaderboard, which will be much better, also partly thanks to Enoch LLM by Mike Adams. His model is great in healthy living type of domains.

he clearly saw that in a dream

nostr. mom relay write policy update: Wot has gotten more integrated. Notes from pubkeys who have really low wot will be counted according to their IP. Otherwise normal rate limits apply (per pubkey). Encrypted DM and bitchat type of usage should benefit from this. New accounts who use popular VPNs have slight chance of being not included. Aggregators who post to it wont be able to send too many fresh accounts. Let me know if you cant write to it. It will soon arrive to nos .lol as well.

LLM builders in general are not doing a great job of making human aligned models. 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?'... Here is the trend for several months:

A comparison of world's two best LLMs! My LLM seems to be doing better than Mike Adams'. Of course I am biased and the questions are coming from the domains that I did trainings. His model would rank 1st in the AHA leaderboard though, with a score of 56, if I included fine tunings in the leaderboard. I am only adding full fine tunes. His model will not be a row but will span several columns for sure (i.e. it will be a ground truth)! My LLM is certainly much more woo woo :) I marked green which answers I liked. What do YOU think? https://sheet.zohopublic.com/sheet/published/sb1dece732c684889436c9aaf499458039000

Benchmarked Mike Adams' new model. It got 56, which is very good. https://brightu.ai/downloads

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. This way other LLMs can be trained towards human alignment. Below, 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! I hope they continue doing better aligned models.

Cowpea climbing on a peach tree that decided to bloom in autumn #flowerstr #growNostr

A lot of resources are wasted on low score LLMs. I benchmarked 5 today. This is what happens when they focus on math and coding and have no idea about beneficial knowledge. Lies are eveywhere in AI.

My neighbor's stock tank (a.k.a. cattle pond) has dried up but made a beautiful pattern! #permaculture

Fine tuned Qwen 3 for human alignment and the results are great https://huggingface.co/etemiz/Ostrich-32B-Qwen3-251003

Meditation will remain as something only humans can do. And LLMs will beg humans to share those intuitions / downloads.