© 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. 22537
    Prev
    Next
    INDEX
    Explanations

    actions related to implementation and development processes

    oai_token-act-pair · gpt-4o-miniTriggered 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
    529
    +0.12
    0.6%
    447
    +0.09
    0.4%
    29
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    529
    +0.12
    0.04
    70
    +0.09
    0.04
    665
    +0.08
    0.04
    Negative Logits
    eria
    -0.82
     proceeding
    -0.69
     awaits
    -0.69
    eq
    -0.69
    journal
    -0.64
    emale
    -0.63
    uyomi
    -0.63
    arat
    -0.63
    tg
    -0.61
     awaiting
    -0.60
    POSITIVE LOGITS
     oneself
    0.80
     them
    0.69
     chunks
    0.63
    ADRA
    0.62
     slices
    0.62
     yourself
    0.62
    redients
    0.61
     keywords
    0.61
    antine
    0.61
    ¬¼
    0.60
    Activations Density 0.179%

    No Known Activations