© 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. 0-RES-JB
    11. 12837
    Prev
    Next
    INDEX
    Explanations

    references to power capacity measured in megawatts

    oai_token-act-pair · gpt-3.5-turbo

    mentions of megawatt (MW) capacity in relation to power generation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Char
    -0.77
     handwriting
    -0.66
     indec
    -0.66
     inscription
    -0.66
     frozen
    -0.65
     passes
    -0.62
     Simon
    -0.62
     alphabet
    -0.62
    Ident
    -0.62
     incons
    -0.61
    POSITIVE LOGITS
    MW
    4.07
     MW
    3.25
    GW
    1.90
     GW
    1.57
    MJ
    1.44
     Mech
    1.40
    NW
    1.36
     LW
    1.34
     kW
    1.24
    CW
    1.18
    Activations Density 0.026%

    No Known Activations