Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 4-RES-JB
    6. 12798
    Prev
    Next
    INDEX
    Explanations

    names of political figures, specifically focusing on the name "Newt Gingrich."

    oai_token-act-pair · gpt-3.5-turbo

    mentions of Newt Gingrich

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 4-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    4
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    fman
    -0.80
    SHIP
    -0.78
     guiActiveUnfocused
    -0.74
    acters
    -0.72
    ournal
    -0.72
    Interstitial
    -0.71
    Production
    -0.70
    hops
    -0.69
    Tokens
    -0.68
    needs
    -0.68
    POSITIVE LOGITS
     Gingrich
    1.42
     Newt
    1.07
     Romney
    0.77
     Benedict
    0.77
     Ruff
    0.77
     Rubio
    0.76
     Giuliani
    0.76
     Cruise
    0.74
    onom
    0.73
    rolet
    0.72
    Activations Density 0.009%

    No Known Activations