Goodfire

AI

sparse autoencoders

SHOW HN2024

201points
51comments
#393of 3,515 in 2024
Top 12%of 2024 by points

How HN took it

CuriousMostly questions and neutral comments

3 positive, 8 neutral or asking, 0 mixed, 4 critical, out of 15 comments about the project. Thread score -0.13.

Idea: 0 for, 2 againstPrivacy: 1 for, 1 againstExecution: 0 for, 1 againstDesign: 1 for, 0 against

The thread, comment by comment

Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.

positive neutral or mixed critical about another commenter or something else the maker

What HN said

What people liked

  • Zoomable UMAP visualisation is nice

    From 1 comment
    • nice work. enjoyed the zoomable UMAP. i wonder if there are hparams to recluster the UMAP in interesting ways. after the idea that Claude 3.5 Sonnet used SAEs to improve its…
      swyx on HN
  • OpenAI-compatible sampling API

    From 1 comment
    • Should we expect the current API (https://docs.goodfire.ai/introduction) to be good for prod or is it for testing only? I see that the sampling API is OpenAI-compatible (nice!).…
      detente18 on HN

Objections

  • Skepticism about generic sparsification efforts

    From 1 comment
    • I am skeptical of generic sparsification efforts. After all, companies like Neural Magic spent years trying to make it work, only to pivot to 'vLLM' engine and be sold to Red Hat
      owenthejumper on HN
  • Login issue: app logs out after first click

    From 1 comment
    • The app keeps logging me out after first click. The tech seems to be intriguiging for me as a software engineer looking to get into custom llm stuff
      Inviz on HN
  • Concern about sending data to a foreign service

    From 1 comment
    • I wonder how many people or companies choose to send their data to foreign services for analysis. Personally, I would approach this with caution and am curious to see how this…
      I_am_tiberius on HN

Requests

  • Native LiteLLM integration with a goodfire provider route

    From 1 comment
    • Should we expect the current API (https://docs.goodfire.ai/introduction) to be good for prod or is it for testing only? I see that the sampling API is OpenAI-compatible (nice!).…
      detente18 on HN
  • Try PaCMAP for the UMAP view

    From 1 comment
    • I'd be really curious to see what happens if you use PaCMAP (https://jmlr.org/papers/volume22/20-1061/20-1061.pdf) and more recent large-scale variants…
      bravura on HN

Questions

  • Is the API meant for production or testing only?

    From 1 comment
    • Should we expect the current API (https://docs.goodfire.ai/introduction) to be good for prod or is it for testing only? I see that the sampling API is OpenAI-compatible (nice!).…
      detente18 on HN
  • Any practical use of SAEs beyond Golden Gate Claude?

    From 1 comment
    • nice work. enjoyed the zoomable UMAP. i wonder if there are hparams to recluster the UMAP in interesting ways. after the idea that Claude 3.5 Sonnet used SAEs to improve its…
      swyx on HN

Alternatives named

  • Anthropic's SAE API

    From 1 comment
    • nice work. enjoyed the zoomable UMAP. i wonder if there are hparams to recluster the UMAP in interesting ways. after the idea that Claude 3.5 Sonnet used SAEs to improve its…
      swyx on HN

Category, keywords, summaries and labels on this page were made by AI (Thread Reader v2) and not checked by hand. Each point links to the comments it came from.

Add a badge to your READMEShown on HackerShow
[![Shown on HackerShow](https://hackershow.org/badge.svg)](https://hackershow.org/show/42495936)