© 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 12288
    9. 8-RES_FS12288-JB
    11. 1658
    Prev
    Next
    INDEX
    Explanations

    written numbers indicating statistics or measurements

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    362
    +0.13
    0.6%
    88
    +0.11
    0.5%
    498
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.13
    0.06
    572
    +0.11
    0.05
    614
    +0.10
    0.04
    Negative Logits
     "$:/
    -0.73
    \<
    -0.73
    hair
    -0.71
    6000
    -0.69
    sonian
    -0.68
    soDeliveryDate
    -0.67
     consulted
    -0.67
    Introduced
    -0.66
    rake
    -0.66
    iHUD
    -0.65
    POSITIVE LOGITS
     starters
    0.96
    bidden
    0.86
    ties
    0.80
    wards
    0.77
    aging
    0.77
    akeru
    0.76
     touchdowns
    0.76
     example
    0.76
    geries
    0.76
     sake
    0.75
    Activations Density 0.195%

    No Known Activations