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. 37132
    Prev
    Next
    INDEX
    Explanations

    references to criminal activity and legal issues

    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
    480
    +0.13
    0.6%
    694
    +0.11
    0.5%
    40
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    711
    +0.13
    0.02
    526
    +0.11
    -0.01
    570
    +0.10
    0.01
    Negative Logits
     Provincial
    -0.66
     partName
    -0.65
    iets
    -0.63
    iple
    -0.62
     Startup
    -0.61
     Patreon
    -0.61
    ishops
    -0.60
     Markets
    -0.60
     Browser
    -0.56
     announcements
    -0.56
    POSITIVE LOGITS
     raping
    0.79
     rapist
    0.77
     genitals
    0.76
     sexually
    0.75
     adultery
    0.72
    "'
    0.71
     vagina
    0.71
     semen
    0.69
    sexual
    0.68
    "?
    0.68
    Activations Density 1.033%

    No Known Activations