© 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. 9872
    Prev
    Next
    INDEX
    Explanations

    statistical data, survey results, and margins of error related to polls and research

    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%
    376
    +0.11
    0.5%
    661
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.02
    661
    +0.11
    0.01
    754
    +0.10
    0.01
    Negative Logits
    ergy
    -0.81
    oxide
    -0.69
    JECT
    -0.68
    spe
    -0.65
    isance
    -0.64
    haus
    -0.64
    alogue
    -0.63
    arus
    -0.63
    uni
    -0.62
     Shiva
    -0.61
    POSITIVE LOGITS
     margin
    1.03
     margins
    1.02
     SHARES
    0.81
    Marg
    0.79
    olicy
    0.79
     separating
    0.74
     favoring
    0.71
    margin
    0.70
     heights
    0.69
    oshenko
    0.68
    Activations Density 0.086%

    No Known Activations