© 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 24576
    9. 8-RES_FS24576-JB
    11. 7161
    Prev
    Next
    INDEX
    Explanations

    emphasized text or formatting

    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
    314
    +0.15
    0.7%
    253
    +0.12
    0.5%
    417
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    253
    +0.15
    0.02
    345
    +0.12
    0.01
    314
    +0.11
    0.00
    Negative Logits
    kson
    -0.69
     misunder
    -0.66
    eness
    -0.66
    adelphia
    -0.64
     provoking
    -0.63
    qqa
    -0.63
    ivated
    -0.63
    ahime
    -0.62
     Leilan
    -0.62
    senal
    -0.62
    POSITIVE LOGITS
    insert
    0.88
    ERROR
    0.87
    blank
    0.84
    TEXT
    0.84
    church
    0.83
    AUT
    0.82
    Insert
    0.81
    âĢ¢âĢ¢
    0.81
    execute
    0.81
    laughs
    0.80
    Activations Density 0.017%

    No Known Activations