© 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. 3639
    Prev
    Next
    INDEX
    Explanations

    legal terms and law-related concepts

    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
    373
    +0.20
    1.0%
    640
    +0.19
    0.9%
    539
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.20
    0.07
    539
    +0.19
    0.05
    373
    +0.09
    -0.01
    Negative Logits
    429
    -0.66
     flaming
    -0.64
     Reincarnated
    -0.62
    cuts
    -0.60
    Ĥ¬
    -0.58
    tin
    -0.57
     Anonymous
    -0.56
     Flame
    -0.56
    Reviewer
    -0.56
    Steam
    -0.56
    POSITIVE LOGITS
    eway
    0.66
    assian
    0.66
     landfall
    0.64
    Bus
    0.64
    eways
    0.62
     dawn
    0.60
    urated
    0.60
    Rh
    0.59
    ubb
    0.59
    ixture
    0.58
    Activations Density 0.144%

    No Known Activations