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 3072
    5. 8-RES_FS3072-JB
    6. 1074
    Prev
    Next
    INDEX
    Explanations

    references to sexual content and behaviors

    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
    384
    +0.11
    0.5%
    288
    +0.10
    0.4%
    1
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.11
    0.07
    682
    +0.10
    0.05
    125
    +0.09
    0.05
    Negative Logits
    Dispatch
    -0.77
    ģĸ
    -0.77
    BLIC
    -0.75
    ĪĴ
    -0.75
    quickShipAvailable
    -0.73
    REC
    -0.72
    IVERS
    -0.71
    Hug
    -0.69
    NPR
    -0.69
    ICA
    -0.68
    POSITIVE LOGITS
     intercourse
    1.02
    Sexual
    0.94
     sex
    0.93
    osexual
    0.93
     steroids
    0.92
     genital
    0.89
     steroid
    0.85
     rape
    0.83
     trafficking
    0.81
     underwear
    0.81
    Activations Density 0.748%

    No Known Activations