© 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. 14842
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal negotiations and agreements

    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
    480
    +0.48
    2.4%
    718
    +0.10
    0.5%
    60
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.48
    0.00
    665
    +0.10
    0.07
    718
    +0.09
    0.07
    Negative Logits
     rejoice
    -0.75
     loves
    -0.68
     Travels
    -0.66
    joice
    -0.65
     beautifully
    -0.65
     wonderfully
    -0.62
     beautiful
    -0.61
     å¤
    -0.60
     lovely
    -0.60
     FANTASY
    -0.59
    POSITIVE LOGITS
     oversight
    0.87
     FOIA
    0.86
     litigation
    0.82
     vetting
    0.79
     jurisd
    0.79
     confidentiality
    0.79
     reimbursement
    0.78
     prosecut
    0.77
     confidential
    0.77
     audit
    0.77
    Activations Density 1.356%

    No Known Activations