mnkv

fundamentally, "linguistic illegibility" is a new term for something that we've known about for about a decade now. In RL the more general ideas is "reward hacking" and in NLP it has been called "semantic drift".

I dislike this term because it doesn't explain where this "illegibility" is coming from. Models are post-trained towards non-linguistic goals with (mostly) non-linguistic rewards. A model's reasoning chain is reinforced if it leads to a correct answer or agentic goal. It doesn't need to be linguistically accurate and meanings can drift over training.

Sounds nice except that these are 1 very small scale model, 1 reranker, and 1 embedding model that are far from frontier LLM level. And they're not open sourced.

As much as I agree with the message, this reads like marketing copy trying to make a big deal out of a tiny model being hosted privately.

you mean like the thinkpad trackpoint?

Just talk to them as if they were already your friend. Most of what you talk about with friends isn't just mutual interests and you start conversations with them all the time.

This blog post describes the basic work of a research engineer and nothing more. The amount of surprise the author has seems to suggest they haven't really worked in ML for very long.

Honestly? This is the best its ever been. Getting stuff to run before huggingface and uv and docker containers with cuda was way worse. Even with full open-source, go try to run a 3+ years old model and codebase. The field just moves very fast.

My ideal is copilot that would evaluate the PR against some basic guidelines that maintainers write down.

And perhaps a way to filter PRs to just contributor PRs would be easy to implement and pretty useful

> the generation of 281,128 augmented examples, from which 1,000 were held out as a benchmark test set.

This model is trained on a custom dataset of 280k examples then tested on 1k very similar examples from the same dataset. Of course it is specialized to outperform general models on this specific task in this specific domain with this specific json format for output.

This is a reasonable hobby project and interesting approach to synthetic data generation but not impressive research.

At minimum you should test your model on other benchmarks that have similar tasks e.g. docbench

reasonable post with a decent analogy explaining on-policy learning, only major thing I take issue with is

> Reinforcement learning is a technical subject—there are whole textbooks written about it.

and then linking to the still wip RLHF book instead of the book on RL: Sutton & Barto.

I think it's pretty obvious it's 1. Given the recent huge, clearly politically-motivated cuts from the current administration, it feels pretty likely that FOIA could be disrupted under the guise of "cost-saving".

And I think you're supposed to be generous to the commenter, not the current administration ;)

how does this compare to zotero?

Good summary of some of the main "theoretical" criticism of LLMs but I feel that it's a bit dated and ignores the recent trend of iterative post-training, especially with human feedback. Major chatbots are no doubt being iteratively refined on the feedback from users i.e. interaction feedback, RLHF, RLAIF. So ChatGPT could fall within the sort of "enactive" perspective on language and definitely goes beyond the issues of static datasets and data completeness.

Sidenote: the authors make a mistake when citing Wittgenstein to find similarity between humans and LLMs. Language modelling on a static dataset is mostly not a language game (see Bender and Koller's section on distributional semantics and caveats on learning meaning from "control codes")

The low quality of this blog post (and all your posts) makes me think you wrote it mostly with AI. The irony is palpable.

AI is clearly enhancing humans in their creation of garbage.