© 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. 20539
    Prev
    Next
    INDEX
    Explanations

    instances of textual criticism or analysis

    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
    480
    +0.33
    1.8%
    680
    +0.11
    0.6%
    718
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    680
    +0.33
    0.05
    480
    +0.11
    -0.03
    718
    +0.08
    0.07
    Negative Logits
    ãĢį
    -0.58
     sinners
    -0.58
     Mankind
    -0.56
    ignty
    -0.55
     virginity
    -0.54
    enery
    -0.52
     raping
    -0.51
    "},"
    -0.51
     KILL
    -0.51
    thood
    -0.50
    POSITIVE LOGITS
     understatement
    0.63
     quotation
    0.61
     sarc
    0.61
     citation
    0.61
     reader
    0.59
     rhetorical
    0.59
     editorial
    0.59
     publication
    0.58
     jargon
    0.58
     commenting
    0.58
    Activations Density 1.907%

    No Known Activations