© 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. 10675
    Prev
    Next
    INDEX
    Explanations

    mentions of scientific research studies published in the Proceedings of the National Academy of Sciences

    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
    394
    +0.12
    0.6%
    404
    +0.09
    0.4%
    256
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    394
    +0.12
    0.02
    97
    +0.09
    0.02
    256
    +0.09
    0.02
    Negative Logits
    behind
    -0.83
    forth
    -0.81
    matter
    -0.69
     Nieto
    -0.69
    gow
    -0.68
    seek
    -0.67
    ahan
    -0.64
    alore
    -0.64
    claw
    -0.64
     shack
    -0.63
    POSITIVE LOGITS
     Sciences
    1.05
     Arts
    0.84
     Pediatrics
    0.82
     Sending
    0.77
     Science
    0.73
    ĵ
    0.70
    Ĵ
    0.70
     Studies
    0.69
    oka
    0.69
    legates
    0.69
    Activations Density 0.051%

    No Known Activations