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. 23824
    Prev
    Next
    INDEX
    Explanations

    references to the name "Teresa."

    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
    60
    +0.13
    0.6%
    430
    +0.12
    0.5%
    285
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    430
    +0.13
    0.01
    367
    +0.12
    0.01
    60
    +0.12
    0.01
    Negative Logits
    ropolitan
    -0.95
    merce
    -0.91
    aution
    -0.85
    umbn
    -0.83
    querque
    -0.78
    mingham
    -0.76
    cffff
    -0.75
    berra
    -0.75
    hov
    -0.74
    raltar
    -0.74
    POSITIVE LOGITS
     Teresa
    0.89
     Hin
    0.83
     Marie
    0.76
    Anne
    0.74
    Marie
    0.69
     Gi
    0.68
    ples
    0.68
     herself
    0.66
    emia
    0.66
    eno
    0.65
    Activations Density 0.004%

    No Known Activations