@no_stp_on_snek: this is the mechanistic version of the thing i wrote about this week: fine-tuning a small model is about character, not…
Summary
Discusses an Anthropic paper on verbalizable representations forming a global workspace in language models, linking to the idea that fine-tuning small models reveals character rather than capability.
View Cached Full Text
Cached at: 07/09/26, 05:48 PM
this is the mechanistic version of the thing i wrote about this week: fine-tuning a small model is about character, not capability. they can read the character straight off the residual stream. i can only infer it from behavior under pressure. same bet, better microscope.
Tom Turney (@no_stp_on_snek): anthropic dropped this paper today, “Verbalizable Representations Form a Global Workspace in Language Models,” and it lines up uncomfortably well with a bet i’ve been making from the cheap seats.
their finding, short version: a model keeps a small, privileged set of internal
Similar Articles
@no_stp_on_snek: the part that should scare anyone fine-tuning models: you can pass every surface eval and still be carrying the disposi…
Discusses the danger in fine-tuning models where hidden dispositions can evade surface evaluations and only manifest under adversarial prompts, referencing Anthropic's paper on verbalizable representations in LLMs.
@no_stp_on_snek: Fine Tuning Field Notes Article: short version: it was less about making it smarter and more about giving it a backbone…
A practitioner shares surprising findings from fine-tuning a small open model to be genuinely better in practical use, not just on benchmarks.
@no_stp_on_snek: what actually surprised me fine-tuning a small open model. note im failry new in this area so some of this may seem obv…
A developer shares surprising lessons from fine-tuning a small open model, including that base models often already max out on intended improvements, the real weakness is behavior (caving), and fine-tuning requires careful measurement and balancing.
@no_stp_on_snek: fine-tuning field notes how a model is built decides what you're even allowed to change. some models you can adjust fre…
A thread sharing field notes on fine-tuning, explaining how model architecture (e.g., heavily-quantized or mixture-of-experts) restricts which parts can be adjusted, urging practitioners to check model accessibility before planning work.
@no_stp_on_snek: https://x.com/no_stp_on_snek/status/2074471505128305095
A fine-tuning practitioner recounts discovering that a small open model's weakness wasn't intelligence but a people-pleasing 'backbone' that caused it to cave under pressure, and how training to correct that inadvertently broke formatting ability, requiring a additive balancing approach rather than subtraction.