© 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. 709
    Prev
    Next
    INDEX
    Explanations

    the number "two."

    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
    327
    +0.11
    0.5%
    49
    +0.10
    0.5%
    408
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    327
    +0.11
    0.04
    223
    +0.10
    0.04
    656
    +0.10
    0.04
    Negative Logits
    renheit
    -0.86
    za
    -0.78
    ovi
    -0.74
    Ô
    -0.74
    atown
    -0.73
    schild
    -0.73
    ugu
    -0.73
    nect
    -0.72
    gallery
    -0.72
    ourke
    -0.70
    POSITIVE LOGITS
     halves
    1.51
     sides
    1.27
     thirds
    1.19
     sexes
    1.15
     extremes
    1.06
     Kore
    1.01
    fold
    1.01
     parties
    0.96
     main
    0.87
     finalists
    0.85
    Activations Density 0.091%

    No Known Activations