© 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. 20064
    Prev
    Next
    INDEX
    Explanations

    quantitative information and statistics

    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
    480
    +0.19
    0.9%
    330
    +0.13
    0.6%
    642
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.19
    0.07
    362
    +0.13
    0.06
    330
    +0.11
    0.04
    Negative Logits
     beforehand
    -0.79
     outwe
    -0.75
     depended
    -0.71
    anooga
    -0.70
    witch
    -0.67
     existed
    -0.65
    iots
    -0.64
     meant
    -0.64
     Organisation
    -0.64
     hadn
    -0.63
    POSITIVE LOGITS
     narrator
    0.70
     POV
    0.69
    veland
    0.66
    ansky
    0.64
    undrum
    0.62
     chronological
    0.62
     memoir
    0.61
     Historic
    0.61
     sqor
    0.60
    xtap
    0.60
    Activations Density 0.700%

    No Known Activations