© 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. 3955
    Prev
    Next
    INDEX
    Explanations

    references to exclusivity or unique access

    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
    4
    +0.10
    0.5%
    585
    +0.10
    0.4%
    656
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    427
    +0.10
    0.01
    579
    +0.10
    0.01
    4
    +0.10
    0.01
    Negative Logits
    ;;;;;;;;;;;;
    -0.82
    olesterol
    -0.78
    apers
    -0.77
    ŃĶ
    -0.74
    enegger
    -0.74
    abases
    -0.74
    DAY
    -0.73
    Introduced
    -0.72
    hematic
    -0.72
    aptic
    -0.72
    POSITIVE LOGITS
     privileges
    0.86
     club
    0.76
     bidder
    0.69
     rights
    0.68
     access
    0.67
    ities
    0.67
     exclus
    0.66
    judicial
    0.65
     exclusively
    0.65
     licensee
    0.64
    Activations Density 0.014%

    No Known Activations