© 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. 18318
    Prev
    Next
    INDEX
    Explanations

    financial figures in millions and multi-year values

    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
    462
    +0.10
    0.4%
    485
    +0.10
    0.4%
    742
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    570
    +0.10
    0.03
    310
    +0.10
    0.05
    319
    +0.10
    0.04
    Negative Logits
    values
    -0.74
    Commerce
    -0.74
    Synopsis
    -0.71
    Thousands
    -0.66
    mediated
    -0.63
    worldly
    -0.63
    hei
    -0.63
    ================================================================
    -0.62
    world
    -0.62
    iannopoulos
    -0.61
    POSITIVE LOGITS
     increments
    1.05
     apiece
    0.83
     deducted
    0.83
     additional
    0.71
     subtract
    0.70
     expended
    0.70
     undecided
    0.68
     toggle
    0.68
     nominal
    0.67
     increment
    0.67
    Activations Density 0.261%

    No Known Activations