© 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. 651
    Prev
    Next
    INDEX
    Explanations

    mentions of specific movie studios, particularly Warner Bros

    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
    475
    +0.14
    0.6%
    99
    +0.12
    0.6%
    288
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    475
    +0.14
    0.02
    99
    +0.12
    0.02
    288
    +0.12
    0.02
    Negative Logits
    omorphic
    -0.83
    icles
    -0.77
    unin
    -0.74
    ingu
    -0.73
    oral
    -0.73
    ihu
    -0.72
    odynam
    -0.72
    ococ
    -0.72
    EStream
    -0.71
    skirts
    -0.71
    POSITIVE LOGITS
     Bros
    1.40
     Brothers
    1.16
     Cable
    0.92
    riors
    0.79
     Warner
    0.73
    rior
    0.73
    sonian
    0.73
    vier
    0.73
     Pictures
    0.70
     Brother
    0.70
    Activations Density 0.091%

    No Known Activations