© 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. 37488
    Prev
    Next
    INDEX
    Explanations

    references to the Milky Way galaxy

    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
    288
    +0.14
    0.6%
    547
    +0.11
    0.5%
    578
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    547
    +0.14
    0.00
    539
    +0.11
    0.00
    578
    +0.10
    0.00
    Negative Logits
    */(
    -0.75
    govtrack
    -0.72
    ually
    -0.69
    idency
    -0.68
    uberty
    -0.68
    xual
    -0.67
    thodox
    -0.67
    orney
    -0.66
    psons
    -0.65
    elvet
    -0.65
    POSITIVE LOGITS
    beans
    0.82
     Galaxy
    0.82
    ota
    0.82
     Ways
    0.77
     Kingdom
    0.76
     Beans
    0.76
    bean
    0.74
    moon
    0.72
     Way
    0.72
     Scrolls
    0.72
    Activations Density 0.003%

    No Known Activations