© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 14581
    Prev
    Next
    INDEX
    Explanations

    numbers related to modifications (e.g., lines added, lines removed) in code repositories

    oai_token-act-pair · gpt-3.5-turboTriggered 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
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    171
    +0.10
    0.5%
    588
    +0.10
    0.4%
    292
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    384
    +0.10
    0.04
    588
    +0.10
    0.04
    500
    +0.09
    0.04
    Negative Logits
     Mech
    -0.72
     Amer
    -0.70
     Bos
    -0.68
     Witches
    -0.68
     Caucasian
    -0.67
     Arabian
    -0.66
     Asian
    -0.65
     Occupations
    -0.65
     Jesuit
    -0.65
    ami
    -0.64
    POSITIVE LOGITS
    destroy
    0.99
    redo
    0.96
     abst
    0.91
     withdrawn
    0.88
     withdraw
    0.86
     diseng
    0.86
    exit
    0.81
     subtract
    0.81
     cance
    0.81
    ancel
    0.80
    Activations Density 0.412%

    No Known Activations