© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 8-RES-JB
    11. 14177
    Prev
    Next
    INDEX
    Explanations

    related content at the end of articles or sections, likely for suggested reading or additional information

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "Related" followed by numbers, suggesting a focus on connections to additional content or references

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Negative Logits
    »Ĵ
    -0.87
    ainted
    -0.81
    UGE
    -0.80
    Ĥİ
    -0.71
    adra
    -0.65
    elfth
    -0.64
    aper
    -0.64
    atur
    -0.63
    IVER
    -0.62
    esan
    -0.62
    POSITIVE LOGITS
     Stories
    1.14
     Content
    0.91
     Links
    0.89
     Articles
    0.88
     Topics
    0.87
     articles
    0.83
     stories
    0.81
     VIDEOS
    0.79
     Coverage
    0.79
     Advertisement
    0.78
    Activations Density 0.028%

    No Known Activations