© 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. 2048
    Prev
    Next
    INDEX
    Explanations

    warnings or disclaimers in texts

    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
    373
    +0.11
    0.5%
    616
    +0.11
    0.5%
    573
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    313
    +0.11
    0.02
    530
    +0.11
    0.01
    396
    +0.10
    0.01
    Negative Logits
    issance
    -0.72
    itage
    -0.72
    sonian
    -0.72
    arters
    -0.69
    externalActionCode
    -0.69
    Engineers
    -0.68
    united
    -0.65
    kefeller
    -0.65
    ths
    -0.63
     mail
    -0.63
    POSITIVE LOGITS
     spoilers
    1.22
    WARNING
    1.01
     Contains
    1.01
    OIL
    1.00
    Warning
    0.92
     WARNING
    0.86
     spoiler
    0.79
     nudity
    0.79
     contains
    0.76
    *****
    0.74
    Activations Density 0.092%

    No Known Activations