© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 3043
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal or regulatory issues

    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
    608
    +0.16
    0.7%
    481
    +0.13
    0.6%
    742
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.16
    0.03
    766
    +0.13
    0.02
    69
    +0.11
    0.02
    Negative Logits
    soever
    -0.93
    mare
    -0.85
    mask
    -0.82
    spawn
    -0.78
    iott
    -0.77
    MAS
    -0.72
    mix
    -0.72
    externalActionCode
    -0.72
    mire
    -0.71
    ranged
    -0.71
    POSITIVE LOGITS
     similarities
    1.02
     anecdotal
    0.85
     inconsistencies
    0.83
     examples
    0.82
     successes
    0.81
     positives
    0.76
     parallels
    0.74
     preced
    0.73
     accomplishments
    0.73
     similarity
    0.72
    Activations Density 0.077%

    No Known Activations