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 24576
    5. 8-RES_FS24576-JB
    6. 23829
    Prev
    Next
    INDEX
    Explanations

    verbs related to actions or processes, particularly those associated with legal or criminal contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    677
    +0.12
    0.5%
    584
    +0.09
    0.4%
    714
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    677
    +0.12
    0.04
    540
    +0.09
    0.05
    665
    +0.09
    0.04
    Negative Logits
    talk
    -0.70
     passage
    -0.69
     Flavoring
    -0.64
    manac
    -0.61
    inders
    -0.59
    render
    -0.58
    ifted
    -0.58
    iew
    -0.58
    ibles
    -0.58
     labels
    -0.56
    POSITIVE LOGITS
     concurrently
    1.06
     jointly
    0.94
     by
    0.92
     differently
    0.89
     indoors
    0.88
     theat
    0.86
     BY
    0.85
    pez
    0.85
     anew
    0.84
     collabor
    0.80
    Activations Density 0.138%

    No Known Activations