© 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. 3-RES-JB
    11. 4249
    Prev
    Next
    INDEX
    Explanations

    text related to illegal/controversial content such as child pornography

    oai_token-act-pair · gpt-3.5-turbo

    references to child pornography

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    erd
    -0.86
    cho
    -0.72
    姫
    -0.72
    ¯¯¯¯¯¯¯¯
    -0.71
    string
    -0.69
    bern
    -0.69
    ees
    -0.69
    sole
    -0.69
    eem
    -0.68
    gs
    -0.68
    POSITIVE LOGITS
     pornography
    1.00
     pornographic
    0.85
     imagery
    0.85
     IMAGES
    0.83
     PHOTO
    0.69
     advertisements
    0.69
    ographers
    0.68
     exhibition
    0.68
    OGR
    0.67
     depicting
    0.67
    Activations Density 0.011%

    No Known Activations