© 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. 797
    Prev
    Next
    INDEX
    Explanations

    actions related to support, help, or influence in a beneficial way

    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
    414
    +0.11
    0.5%
    101
    +0.10
    0.5%
    643
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.11
    0.05
    665
    +0.10
    0.04
    643
    +0.10
    0.02
    Negative Logits
    NetMessage
    -0.67
    lio
    -0.64
    llo
    -0.63
    arij
    -0.63
    ModLoader
    -0.62
     razor
    -0.59
     Phant
    -0.56
    ogl
    -0.56
    Ħ¢
    -0.56
     Tsukuyomi
    -0.56
    POSITIVE LOGITS
     perpetuate
    0.82
    spread
    0.82
     amplify
    0.77
     alleviate
    0.76
     legitim
    0.76
     facilitate
    0.74
    abet
    0.74
     perpet
    0.73
    udeb
    0.72
     coordinate
    0.71
    Activations Density 0.234%

    No Known Activations