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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & 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. 46151
    Prev
    Next
    INDEX
    Explanations

    references to copyright information

    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
    609
    +0.14
    0.7%
    151
    +0.13
    0.6%
    214
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    609
    +0.14
    0.02
    151
    +0.13
    0.02
    214
    +0.10
    0.01
    Negative Logits
     Haku
    -0.72
    stones
    -0.62
     Peb
    -0.62
    naires
    -0.62
    ordinary
    -0.61
    abouts
    -0.60
     Lans
    -0.60
    oward
    -0.60
    cki
    -0.59
     cav
    -0.59
    POSITIVE LOGITS
    Copyright
    1.10
     ©
    0.98
    ©
    0.88
     infringement
    0.82
     Copyright
    0.80
    yright
    0.76
     copyright
    0.76
    !/
    0.71
     Provided
    0.69
     Associated
    0.65
    Activations Density 0.009%

    No Known Activations