© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 6922
    Prev
    Next
    INDEX
    Explanations

    proper nouns and titles

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    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
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    480
    +0.21
    1.0%
    87
    +0.12
    0.6%
    640
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    632
    +0.21
    0.01
    764
    +0.12
    0.01
    64
    +0.11
    -0.01
    Negative Logits
    schild
    -0.73
     vanity
    -0.66
    bies
    -0.66
     footing
    -0.64
     fronts
    -0.63
     fert
    -0.63
    bour
    -0.63
     treasury
    -0.62
     bait
    -0.62
     charm
    -0.61
    POSITIVE LOGITS
     ILCS
    1.13
     Immun
    0.76
     largeDownload
    0.76
    NetMessage
    0.75
     =>
    0.72
     DISTRICT
    0.71
     Frames
    0.71
     Unloaded
    0.71
     SHARES
    0.70
    ETF
    0.70
    Activations Density 0.021%

    No Known Activations