© 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. 3228
    Prev
    Next
    INDEX
    Explanations

    email correspondence within the context of medical research, healthcare discussions, and governmental interactions

    oai_token-act-pair · gpt-3.5-turbo
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    314
    +0.14
    0.7%
    91
    +0.10
    0.5%
    4
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    547
    +0.14
    0.04
    417
    +0.10
    0.04
    317
    +0.09
    0.03
    Negative Logits
     idols
    -0.71
    Tags
    -0.70
    rities
    -0.68
    roo
    -0.67
    otin
    -0.63
    Os
    -0.60
     projectiles
    -0.60
    acists
    -0.59
    CHAPTER
    -0.58
    Mods
    -0.56
    POSITIVE LOGITS
     emailed
    0.94
     statement
    0.84
     accompanying
    0.79
     obtained
    0.75
    mail
    0.74
     newsletter
    0.71
    statement
    0.71
    tained
    0.71
    velop
    0.70
    edin
    0.70
    Activations Density 0.085%

    No Known Activations