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 3072
    5. 8-RES_FS3072-JB
    6. 2738
    Prev
    Next
    INDEX
    Explanations

    adverbs or phrases indicating additional information or progression

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    459
    +0.15
    0.7%
    460
    +0.11
    0.5%
    402
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    221
    +0.15
    0.04
    402
    +0.11
    0.04
    455
    +0.10
    0.04
    Negative Logits
    ldom
    -0.73
    liest
    -0.73
    cius
    -0.71
    odor
    -0.71
    ipe
    -0.67
    ertodd
    -0.67
    pei
    -0.66
     catcher
    -0.66
    orno
    -0.63
    essee
    -0.61
    POSITIVE LOGITS
     ado
    0.96
    ance
    0.94
     refinement
    0.89
     subdiv
    0.80
     avenues
    0.77
     developments
    0.77
     compl
    0.77
     installments
    0.76
     exploration
    0.75
     advances
    0.75
    Activations Density 0.322%

    No Known Activations