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. 11520
    Prev
    Next
    INDEX
    Explanations

    legal titles and regulations, specifically related to gender identity and anti-discrimination laws

    oai_token-act-pair · gpt-3.5-turboTriggered 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
    167
    +0.13
    0.6%
    164
    +0.12
    0.6%
    316
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    164
    +0.13
    0.01
    437
    +0.12
    0.01
    214
    +0.10
    0.01
    Negative Logits
     whisk
    -0.69
    eu
    -0.62
     detachment
    -0.62
    Äĩ
    -0.62
    EMS
    -0.61
    addafi
    -0.61
     virt
    -0.60
     Leone
    -0.60
     Canaver
    -0.58
     Spice
    -0.58
    POSITIVE LOGITS
    \\\\\\\\
    0.92
     IX
    0.90
    aho
    0.88
    istries
    0.79
     VII
    0.78
    ĨĴ
    0.78
    itional
    0.77
    nance
    0.74
    ewater
    0.74
    payer
    0.73
    Activations Density 0.009%

    No Known Activations