© 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. 1702
    Prev
    Next
    INDEX
    Explanations

    violent and harmful actions or events such as beheadings, stranglings, and decapitations

    oai_token-act-pair · gpt-3.5-turbo

    words related to violence and death

    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
    WE
    -0.75
    METHOD
    -0.66
     OB
    -0.66
    FW
    -0.66
    uph
    -0.65
    WB
    -0.64
    Stre
    -0.64
    TERN
    -0.64
    forward
    -0.62
    BN
    -0.62
    POSITIVE LOGITS
     Victims
    0.74
     overboard
    0.72
     drowned
    0.72
     confisc
    0.71
     kidnap
    0.71
    bage
    0.71
    nces
    0.68
     runaway
    0.67
    geoning
    0.66
     assassinated
    0.66
    Activations Density 0.169%

    No Known Activations