© 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. 3-RES-JB
    11. 4601
    Prev
    Next
    INDEX
    Explanations

    technology-related terms, specifically related to the Mozilla organization

    oai_token-act-pair · gpt-3.5-turbo

    references to the Mozilla organization

    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
    Interstitial
    -0.86
    draw
    -0.78
    indu
    -0.70
    ivism
    -0.70
    sing
    -0.68
     Qian
    -0.68
    haar
    -0.67
    payer
    -0.67
    lie
    -0.67
    mix
    -0.66
    POSITIVE LOGITS
     Firefox
    0.83
    isoft
    0.81
     Mozilla
    0.76
    auga
    0.75
    conservancy
    0.72
     mosqu
    0.72
    avascript
    0.71
    ĺħ
    0.69
     handshake
    0.69
     interoper
    0.68
    Activations Density 0.007%

    No Known Activations