© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 24174
    Prev
    Next
    INDEX
    Explanations

    phrases that include prepositions followed by nouns associated with actions or decisions

    oai_token-act-pair · gpt-3.5-turbo

    phrases highlighting actions and statements that are controversial or involve political discourse

    oai_token-act-pair · gpt-4o-miniTriggered 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
    Not in Any Lists

    No Comments

    Negative Logits
    fml
    -0.83
    /"
    -0.74
    oreal
    -0.73
    forge
    -0.71
    MpServer
    -0.69
     resil
    -0.68
    aves
    -0.65
    ibble
    -0.65
    =#
    -0.63
    fixed
    -0.62
    POSITIVE LOGITS
     politicians
    0.93
     individuals
    0.82
     successive
    0.80
     Senators
    0.79
     those
    0.79
     Senator
    0.77
     pundits
    0.76
     businessmen
    0.76
     Presidents
    0.75
     outsiders
    0.74
    Activations Density 0.247%

    No Known Activations