© 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. 3706
    Prev
    Next
    INDEX
    Explanations

    web-related terms and technology updates

    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
    +0.16
    0.7%
    327
    +0.12
    0.6%
    720
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    720
    +0.16
    0.08
    327
    +0.12
    0.08
    675
    +0.12
    0.07
    Negative Logits
     looms
    -0.67
     fortun
    -0.66
     loopholes
    -0.62
     smugglers
    -0.61
     Clever
    -0.59
     decimal
    -0.59
     Mush
    -0.59
     Canaan
    -0.58
     miscar
    -0.58
     silhou
    -0.58
    POSITIVE LOGITS
    iversary
    0.90
    ï¸ı
    0.85
    office
    0.82
    same
    0.81
    ski
    0.79
    ution
    0.79
    tyard
    0.78
    aut
    0.77
    ship
    0.77
    worthiness
    0.76
    Activations Density 0.162%

    No Known Activations