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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsBlogSlackPrivacy & 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. 40054
    Prev
    Next
    INDEX
    Explanations

    instances of the word "smack" and its variations, focusing on metaphors of impact or criticism

    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
    526
    +0.13
    0.6%
    636
    +0.11
    0.5%
    316
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    526
    +0.13
    0.00
    636
    +0.11
    0.01
    42
    +0.10
    0.01
    Negative Logits
     Seasons
    -0.75
    egal
    -0.74
    abama
    -0.71
    efficient
    -0.70
    efficiency
    -0.70
    psons
    -0.66
    pmwiki
    -0.65
     Lynd
    -0.65
    emale
    -0.64
    MSN
    -0.64
    POSITIVE LOGITS
     dab
    1.56
    down
    1.13
     smack
    1.06
    downs
    0.93
    stick
    0.92
     bang
    0.90
    dash
    0.90
    boxing
    0.85
    bowl
    0.83
    manship
    0.80
    Activations Density 0.006%

    No Known Activations