© 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. 19918
    Prev
    Next
    INDEX
    Explanations

    references to television programs and related terms

    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
    360
    +0.11
    0.5%
    493
    +0.11
    0.5%
    167
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    493
    +0.11
    0.03
    360
    +0.11
    0.03
    725
    +0.10
    0.03
    Negative Logits
    xual
    -0.86
    forth
    -0.79
    cies
    -0.78
    ussen
    -0.78
    20439
    -0.73
    iveness
    -0.73
    ebook
    -0.71
    currency
    -0.71
    cy
    -0.70
    etsy
    -0.67
    POSITIVE LOGITS
     Trop
    1.10
    trop
    1.05
     Shows
    0.95
     commercials
    0.88
    ILLE
    0.88
     presenter
    0.87
     broadcasts
    0.85
    caster
    0.81
    NZ
    0.81
     viewership
    0.80
    Activations Density 0.028%

    No Known Activations