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

    Neuronpedia

    Jacobian LensNEW
    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. 17805
    Prev
    Next
    INDEX
    Explanations

    information related to additional details reported in the text

    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
    60
    +0.13
    0.6%
    402
    +0.10
    0.5%
    623
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    402
    +0.13
    0.02
    60
    +0.10
    0.02
    623
    +0.10
    0.02
    Negative Logits
    =-=-=-=-=-=-=-=-
    -0.71
    generic
    -0.69
    lua
    -0.67
     Gone
    -0.67
     Alone
    -0.66
    FORE
    -0.65
    athon
    -0.64
    ÅĤ
    -0.64
    rpm
    -0.63
    debian
    -0.63
    POSITIVE LOGITS
    judicial
    1.07
     impetus
    0.96
     layers
    0.95
     installments
    0.91
     avenues
    0.83
     complication
    0.82
    ordinary
    0.82
     dimensions
    0.81
     scrutiny
    0.80
     information
    0.79
    Activations Density 0.032%

    No Known Activations