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. 23861
    Prev
    Next
    INDEX
    Explanations

    instances of the word "that" in various grammatical contexts

    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
    393
    +0.15
    0.7%
    60
    +0.13
    0.6%
    97
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    60
    +0.15
    0.03
    655
    +0.13
    0.03
    109
    +0.11
    0.03
    Negative Logits
     Sheet
    -0.70
     Ruk
    -0.66
    apsed
    -0.65
     downed
    -0.65
    PDATE
    -0.60
    ono
    -0.60
     dying
    -0.60
     Sitting
    -0.58
     TheNitromeFan
    -0.57
    ahime
    -0.57
    POSITIVE LOGITS
     resembles
    1.01
     preserves
    0.90
     mirrors
    0.89
     minimize
    0.88
     avoids
    0.87
     maximize
    0.86
     reflects
    0.86
     violates
    0.85
     respects
    0.85
     allows
    0.85
    Activations Density 0.066%

    No Known Activations