© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 19593
    Prev
    Next
    INDEX
    Explanations

    derogatory and offensive language

    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
    481
    +0.12
    0.6%
    91
    +0.11
    0.5%
    577
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    32
    +0.12
    0.04
    568
    +0.11
    0.03
    91
    +0.10
    0.04
    Negative Logits
    enhagen
    -0.72
    escal
    -0.68
     corrid
    -0.66
    earchers
    -0.66
    ãĥ¼ãĥ³
    -0.65
    paren
    -0.64
    restricted
    -0.64
     Passage
    -0.64
     bilateral
    -0.63
     uninterrupted
    -0.63
    POSITIVE LOGITS
    fuck
    1.06
     bastard
    1.01
     bitch
    0.96
     asshole
    0.96
    hole
    0.95
    gery
    0.95
     hypocr
    0.94
     cunt
    0.94
     crap
    0.91
     liar
    0.89
    Activations Density 0.167%

    No Known Activations