© 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. 14821
    Prev
    Next
    INDEX
    Explanations

    medical terms or terms related to healthcare and procedures

    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
    87
    +0.15
    0.7%
    439
    +0.11
    0.5%
    240
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    627
    +0.15
    0.03
    87
    +0.11
    -0.04
    15
    +0.10
    0.06
    Negative Logits
    Conclusion
    -0.92
    )).
    -0.83
    CONCLUS
    -0.81
     UNCLASSIFIED
    -0.77
    âĶĢâĶĢâĶĢâĶĢ
    -0.73
    endif
    -0.71
    .).
    -0.69
     blat
    -0.69
    ))))
    -0.69
    })
    -0.69
    POSITIVE LOGITS
     Thursday
    0.77
     Wednesday
    0.75
     Tuesday
    0.74
     Monday
    0.70
     Friday
    0.69
     reportedly
    0.66
     unveiled
    0.64
     lodged
    0.60
     announced
    0.58
     posted
    0.58
    Activations Density 1.235%

    No Known Activations