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. 25577
    Prev
    Next
    INDEX
    Explanations

    the transitional phrase "Anyway" and variations of it

    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
    709
    +0.12
    0.5%
    55
    +0.10
    0.5%
    415
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    501
    +0.12
    0.01
    267
    +0.10
    0.01
    450
    +0.10
    0.01
    Negative Logits
    adr
    -0.62
    prus
    -0.59
    "},"
    -0.58
     basket
    -0.56
     solitary
    -0.56
     prost
    -0.55
    rament
    -0.55
    ream
    -0.54
    lé
    -0.52
     centerpiece
    -0.52
    POSITIVE LOGITS
    adays
    0.95
     anyways
    0.86
     congratulations
    0.84
     suffice
    0.81
     congr
    0.81
     Anyway
    0.76
     ado
    0.75
    Anyway
    0.73
     suppose
    0.71
    herty
    0.71
    Activations Density 0.008%

    No Known Activations