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

    Neuronpedia

    APISteerSAE EvalsBlog/PodcastNEWSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlog/PodcastGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 3072
    5. 8-RES_FS3072-JB
    6. 1207
    Prev
    Next
    INDEX
    Explanations

    mentions of a specific female individual

    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
    612
    +0.15
    0.7%
    610
    +0.11
    0.5%
    526
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    612
    +0.15
    0.11
    610
    +0.11
    0.08
    15
    +0.10
    0.09
    Negative Logits
    kefeller
    -0.85
    ype
    -0.83
    ypes
    -0.81
    vernment
    -0.74
    undo
    -0.73
    ornia
    -0.72
    antage
    -0.66
    ustom
    -0.65
    hovah
    -0.64
    redo
    -0.64
    POSITIVE LOGITS
    pher
    1.38
     herself
    1.37
    athed
    1.20
     husband
    1.13
    ding
    1.13
    athing
    1.11
    pard
    1.11
    metic
    1.04
    cule
    1.04
     vagina
    0.96
    Activations Density 0.533%

    No Known Activations