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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 12288
    9. 8-RES_FS12288-JB
    11. 885
    Prev
    Next
    INDEX
    Explanations

    adjectives describing size, importance, or impact

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    718
    +0.10
    0.5%
    697
    +0.10
    0.5%
    362
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    65
    +0.10
    0.07
    362
    +0.10
    0.08
    408
    +0.10
    0.09
    Negative Logits
    events
    -0.92
    hops
    -0.90
    tags
    -0.88
    iddles
    -0.86
    speak
    -0.86
     balloons
    -0.79
    hots
    -0.78
    inances
    -0.77
    laws
    -0.77
    casts
    -0.77
    POSITIVE LOGITS
     amount
    1.21
     swath
    1.19
     assortment
    1.09
     subset
    1.06
     chunk
    1.06
     dose
    1.05
     handful
    1.03
     uptick
    1.01
     degree
    1.00
     bunch
    0.96
    Activations Density 0.541%

    No Known Activations