© 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 EvalsExportsinterp-engineNEWAPI 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. 24422
    Prev
    Next
    INDEX
    Explanations

    phrases indicating a list or examples

    oai_token-act-pair · gpt-4o-miniTriggered 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
    39
    +0.11
    0.5%
    220
    +0.10
    0.4%
    481
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.11
    0.04
    276
    +0.10
    0.03
    622
    +0.09
    0.03
    Negative Logits
    akery
    -0.86
    uca
    -0.77
    undle
    -0.77
    livion
    -0.76
    iosity
    -0.75
    amphetamine
    -0.72
    fecture
    -0.72
    etz
    -0.69
    ardless
    -0.69
    opus
    -0.68
    POSITIVE LOGITS
     examples
    1.30
     highlights
    1.21
     excerpts
    1.12
     testim
    1.05
     noteworthy
    1.03
     notable
    1.03
     pointers
    1.02
     illust
    1.02
     possibilities
    0.99
     suggestions
    0.97
    Activations Density 0.113%

    No Known Activations