© 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. 20067
    Prev
    Next
    INDEX
    Explanations

    special characters and non-English characters

    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
    350
    +0.13
    0.6%
    40
    +0.11
    0.5%
    304
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    350
    +0.13
    0.03
    639
    +0.11
    0.03
    398
    +0.10
    0.03
    Negative Logits
    inav
    -0.77
    channelAvailability
    -0.76
    azeera
    -0.74
    hani
    -0.73
    cffffcc
    -0.72
    anova
    -0.68
    halla
    -0.68
    ulhu
    -0.68
    ounty
    -0.67
    isSpecialOrderable
    -0.67
    POSITIVE LOGITS
    redo
    0.65
    ëĭ
    0.62
    ilon
    0.61
     Romans
    0.60
    ÙĴ
    0.58
    hoff
    0.58
    ÃŁ
    0.58
    åĩ
    0.56
    DonaldTrump
    0.56
    cigarette
    0.55
    Activations Density 0.133%

    No Known Activations