© 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. 3080
    Prev
    Next
    INDEX
    Explanations

    prepositions and helping verbs

    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
    608
    +0.17
    0.8%
    718
    +0.11
    0.5%
    762
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.17
    0.03
    718
    +0.11
    0.03
    49
    +0.10
    0.02
    Negative Logits
    yet
    -0.75
    bats
    -0.70
    Reviewed
    -0.66
    hello
    -0.65
    quished
    -0.65
    lings
    -0.65
    plate
    -0.65
    calling
    -0.64
     Closing
    -0.63
    door
    -0.63
    POSITIVE LOGITS
     prioritize
    1.16
     accumulate
    1.15
     behave
    1.09
     attract
    1.07
     stick
    1.06
     specialize
    1.05
     forget
    1.04
     underestimate
    1.03
     adhere
    1.02
     settle
    1.02
    Activations Density 0.058%

    No Known Activations