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. 43383
    Prev
    Next
    INDEX
    Explanations

    reference to publication details and locations

    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
    373
    +0.16
    0.7%
    673
    +0.09
    0.4%
    579
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    673
    +0.16
    0.01
    373
    +0.09
    -0.00
    308
    +0.09
    0.01
    Negative Logits
    SPONSORED
    -0.85
    iqueness
    -0.78
     plot
    -0.68
    xual
    -0.67
     tripod
    -0.66
    vati
    -0.66
     plots
    -0.66
     extras
    -0.66
    iru
    -0.65
    halla
    -0.65
    POSITIVE LOGITS
     Springer
    0.89
     Gale
    0.87
     Symphony
    0.83
     Univ
    0.82
     Journal
    0.80
     Wiley
    0.79
     Brill
    0.78
     Printed
    0.76
     Zed
    0.76
     Jos
    0.75
    Activations Density 0.016%

    No Known Activations