© 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 24576
    9. 8-RES_FS24576-JB
    11. 18560
    Prev
    Next
    INDEX
    Explanations

    parenthetical comments at the end of sentences

    oai_token-act-pair · gpt-3.5-turboTriggered 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
    640
    +0.19
    0.9%
    496
    +0.14
    0.6%
    6
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    6
    +0.19
    0.04
    71
    +0.14
    0.03
    636
    +0.12
    0.03
    Negative Logits
     inund
    -0.82
     undet
    -0.80
     stagn
    -0.78
     spir
    -0.77
     footing
    -0.74
     pudding
    -0.74
    ogly
    -0.74
     exemplary
    -0.74
     overwhelmed
    -0.73
     unus
    -0.73
    POSITIVE LOGITS
    âĢ¦)
    1.38
    Though
    1.25
    Laughs
    1.25
    See
    1.23
    laughs
    1.22
    Unless
    1.22
    Ironically
    1.20
    Also
    1.20
    hide
    1.19
    Actually
    1.17
    Activations Density 0.037%

    No Known Activations