© 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. 3287
    Prev
    Next
    INDEX
    Explanations

    personal stories and experiences, particularly related to challenges and hardships

    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
    +0.17
    0.8%
    586
    +0.10
    0.5%
    23
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    87
    +0.17
    -0.04
    586
    +0.10
    0.03
    468
    +0.09
    0.04
    Negative Logits
    Were
    -0.62
     Were
    -0.62
    """
    -0.61
    were
    -0.61
     *)
    -0.60
    iots
    -0.58
     ourselves
    -0.58
    biased
    -0.56
     Inquiry
    -0.55
     Detected
    -0.55
    POSITIVE LOGITS
     his
    0.82
     himself
    0.78
     tattoos
    0.75
     herself
    0.71
     scars
    0.65
     autobiography
    0.65
    His
    0.64
    his
    0.63
     pals
    0.62
     tattoo
    0.62
    Activations Density 1.358%

    No Known Activations