© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 8-RES-JB
    11. 13004
    Prev
    Next
    INDEX
    Explanations

    acronyms and technical specifications

    oai_token-act-pair · gpt-3.5-turbo

    alphanumeric codes or identifiers

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Negative Logits
     Nirvana
    -0.72
     Bulgar
    -0.65
     mall
    -0.65
     Diablo
    -0.63
     Chevy
    -0.63
     Juda
    -0.62
     Giuliani
    -0.61
     Bethesda
    -0.61
     Antar
    -0.60
     Addiction
    -0.60
    POSITIVE LOGITS
    ipped
    1.00
    illed
    0.92
    agged
    0.92
    OUN
    0.91
    ailed
    0.91
    ached
    0.87
    ATE
    0.86
    itted
    0.85
    icked
    0.83
    apped
    0.83
    Activations Density 0.195%

    No Known Activations