© 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. 18760
    Prev
    Next
    INDEX
    Explanations

    references to sports records and technical specifications

    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
    +0.23
    1.1%
    481
    +0.20
    1.0%
    168
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.23
    0.08
    314
    +0.20
    0.00
    627
    +0.12
    0.03
    Negative Logits
     Flavoring
    -0.76
     Rolls
    -0.70
     pressing
    -0.68
     Ferrari
    -0.64
     Glock
    -0.63
     Boeing
    -0.63
     Guant
    -0.63
     rocket
    -0.62
     Austrian
    -0.62
     lesbians
    -0.62
    POSITIVE LOGITS
    CLASS
    0.81
    HQ
    0.73
    pedia
    0.72
    Official
    0.72
    WC
    0.68
    OSE
    0.67
    arus
    0.66
    udos
    0.66
    obal
    0.66
    cn
    0.66
    Activations Density 0.438%

    No Known Activations