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

    Neuronpedia

    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. 16177
    Prev
    Next
    INDEX
    Explanations

    mentions of legal convictions, especially for serious crimes such as felonies, murders, and sex offenses

    oai_token-act-pair · gpt-3.5-turbo

    references to legal convictions

    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
    arity
    -0.71
    edia
    -0.68
     Gork
    -0.67
    ophe
    -0.65
    yip
    -0.64
    abwe
    -0.64
    VEL
    -0.63
    mentation
    -0.63
    earable
    -0.62
    aven
    -0.62
    POSITIVE LOGITS
     felon
    0.94
    icts
    0.84
     convicted
    0.82
     unfocusedRange
    0.80
     guilty
    0.79
     convict
    0.78
     debtor
    0.77
    iary
    0.72
    icted
    0.71
    eering
    0.71
    Activations Density 0.016%

    No Known Activations