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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 12288
    9. 8-RES_FS12288-JB
    11. 9738
    Prev
    Next
    INDEX
    Explanations

    the word "gas" at various intensities

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    288
    +0.13
    0.6%
    314
    +0.11
    0.5%
    12
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    12
    +0.13
    0.03
    288
    +0.11
    0.03
    442
    +0.10
    0.03
    Negative Logits
    lihood
    -0.89
    ership
    -0.83
    enance
    -0.77
    reads
    -0.70
     Bei
    -0.70
     Arbor
    -0.66
    ournal
    -0.63
    âĢ¢âĢ¢
    -0.63
    ctive
    -0.60
     Bald
    -0.60
    POSITIVE LOGITS
    oline
    1.31
    lighting
    0.89
    olina
    0.87
     mileage
    0.86
     liquids
    0.84
    ping
    0.84
     chambers
    0.81
     stations
    0.80
     pedal
    0.80
     turbines
    0.78
    Activations Density 0.067%

    No Known Activations