© 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. 3202
    Prev
    Next
    INDEX
    Explanations

    instances of the word 'however' and phrases indicating that a subject is associated with or defined by something specific

    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
    447
    +0.22
    1.0%
    661
    +0.10
    0.5%
    701
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    661
    +0.22
    0.03
    8
    +0.10
    0.04
    498
    +0.10
    0.04
    Negative Logits
    enced
    -0.70
     Yuan
    -0.59
     Vaughan
    -0.54
    Tam
    -0.53
    nea
    -0.53
     Hat
    -0.53
     Turk
    -0.53
     Heller
    -0.52
     Kaplan
    -0.51
     Oakland
    -0.50
    POSITIVE LOGITS
     guiActiveUn
    0.73
    soever
    0.72
    iton
    0.69
    EVA
    0.68
    pton
    0.67
    vc
    0.67
    unders
    0.67
    FW
    0.66
    tan
    0.66
    =-=-=-=-
    0.66
    Activations Density 0.404%

    No Known Activations