© 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. 15533
    Prev
    Next
    INDEX
    Explanations

    contractions and negations

    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
    551
    +0.14
    0.6%
    360
    +0.10
    0.5%
    383
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    551
    +0.14
    0.07
    139
    +0.10
    0.06
    383
    +0.09
    0.05
    Negative Logits
    itiz
    -0.71
    forcement
    -0.69
     Seasons
    -0.69
     Opportun
    -0.69
    accompan
    -0.69
    ãĤ¼
    -0.67
     Passage
    -0.67
     Laun
    -0.66
    thal
    -0.65
    çĶŁ
    -0.64
    POSITIVE LOGITS
     know
    1.33
     deserve
    1.22
     necessarily
    1.21
     intend
    1.20
     belong
    1.19
     understand
    1.15
     want
    1.14
     bother
    1.13
     exist
    1.12
     seem
    1.12
    Activations Density 0.142%

    No Known Activations