Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 26321
    Prev
    Next
    INDEX
    Explanations

    references to political activities and organizations

    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
    650
    +0.10
    0.5%
    723
    +0.10
    0.5%
    632
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    247
    +0.10
    0.02
    723
    +0.10
    0.01
    110
    +0.10
    0.01
    Negative Logits
     recall
    -0.70
    ged
    -0.69
    ragon
    -0.67
    raped
    -0.65
     KH
    -0.62
    lessly
    -0.62
     redundant
    -0.61
    shire
    -0.60
    aye
    -0.60
    locked
    -0.60
    POSITIVE LOGITS
     Committee
    0.86
     committee
    0.80
     Institute
    0.80
     institute
    0.78
     Conference
    0.72
     Committees
    0.68
     Dinner
    0.67
     Commission
    0.67
     commission
    0.66
     Caucus
    0.65
    Activations Density 0.027%

    No Known Activations