© 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. 15091
    Prev
    Next
    INDEX
    Explanations

    descriptions related to sensory experiences

    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
    547
    +0.11
    0.5%
    644
    +0.10
    0.5%
    89
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    724
    +0.11
    0.05
    225
    +0.10
    0.05
    712
    +0.10
    0.03
    Negative Logits
    king
    -0.69
     papers
    -0.67
     aborted
    -0.63
    Blocks
    -0.63
    References
    -0.62
     cages
    -0.62
     copies
    -0.62
     Articles
    -0.61
     paperwork
    -0.60
    Reviewed
    -0.59
    POSITIVE LOGITS
    tones
    1.10
     perv
    1.09
     reminiscent
    0.98
     vibe
    0.96
    enance
    0.87
    iness
    0.86
     accent
    0.85
    notations
    0.84
     contrasted
    0.84
     sensibilities
    0.84
    Activations Density 0.223%

    No Known Activations