© 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 EvalsExportsAPI 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. 4-RES-JB
    11. 1081
    Prev
    Next
    INDEX
    Explanations

    words related to destructive actions or damage to property

    oai_token-act-pair · gpt-3.5-turbo

    instances of vandalism and related terms

    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
    croft
    -0.84
    ramid
    -0.79
    erald
    -0.75
    hart
    -0.74
    obar
    -0.74
    atches
    -0.73
    nesota
    -0.71
    pres
    -0.69
    rera
    -0.69
    omo
    -0.68
    POSITIVE LOGITS
     vandalism
    1.14
     graffiti
    1.02
     spree
    0.96
     vandal
    0.93
     leve
    0.72
     dere
    0.71
     rampage
    0.69
     detract
    0.69
     inflicted
    0.68
    icons
    0.66
    Activations Density 0.013%

    No Known Activations