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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 49152
    9. 8-RES_FS49152-JB
    11. 25727
    Prev
    Next
    INDEX
    Explanations

    statements and definitions related to rights and regulations concerning individuals and organizations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    256
    +0.14
    0.7%
    481
    +0.12
    0.6%
    263
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    256
    +0.14
    0.04
    263
    +0.12
    0.04
    100
    +0.12
    0.04
    Negative Logits
     stuff
    -0.85
     mug
    -0.78
    Laughs
    -0.77
     Lup
    -0.76
     xp
    -0.73
     crap
    -0.73
     lol
    -0.70
     gigs
    -0.70
    stuff
    -0.70
     shit
    -0.69
    POSITIVE LOGITS
     hereby
    1.43
     herein
    0.98
    '"
    0.98
     unlawfully
    0.95
    withstanding
    0.94
    shall
    0.94
     unlawful
    0.93
     Accordingly
    0.92
     Plaintiff
    0.92
     Plaint
    0.90
    Activations Density 0.289%

    No Known Activations