© 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. 14942
    Prev
    Next
    INDEX
    Explanations

    details about gelatin consumption and penny savings

    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
    465
    +0.10
    0.5%
    580
    +0.09
    0.4%
    571
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    227
    +0.10
    0.03
    30
    +0.09
    0.03
    455
    +0.09
    0.03
    Negative Logits
     Crom
    -0.73
     Schwarz
    -0.68
    retch
    -0.68
     Mahjong
    -0.66
    throp
    -0.65
    endar
    -0.62
     Pok
    -0.62
    etsk
    -0.61
     Tibet
    -0.60
    wise
    -0.60
    POSITIVE LOGITS
    ÃĥÃĤ
    0.79
    uca
    0.76
    Ļ
    0.75
    çĶ°
    0.75
    ãĤ¨ãĥ«
    0.71
    ilyn
    0.67
    Orderable
    0.65
    pen
    0.64
    ãĥĸ
    0.64
    =\"
    0.63
    Activations Density 0.107%

    No Known Activations