© 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 49152
    9. 8-RES_FS49152-JB
    11. 37284
    Prev
    Next
    INDEX
    Explanations

    occurrences of the name "Schultz" in the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    526
    +0.15
    0.7%
    659
    +0.14
    0.6%
    578
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.15
    0.01
    430
    +0.14
    0.01
    111
    +0.13
    0.01
    Negative Logits
    ccording
    -0.76
     corrid
    -0.75
    ãĤ¤ãĥĪ
    -0.74
    66666666
    -0.70
    ãģ¦
    -0.69
    ãĥ¼ãĥ³
    -0.69
    20439
    -0.68
    ����
    -0.66
     Lithuan
    -0.65
    icative
    -0.65
    POSITIVE LOGITS
     Schultz
    0.91
    olicy
    0.90
    midt
    0.84
    meyer
    0.79
    ulz
    0.76
    steen
    0.76
    mann
    0.75
    icago
    0.74
    uler
    0.74
    hoff
    0.73
    Activations Density 0.004%

    No Known Activations