© 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 24576
    9. 8-RES_FS24576-JB
    11. 18850
    Prev
    Next
    INDEX
    Explanations

    information about the origin or source of different entities

    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
    480
    +0.18
    0.8%
    399
    +0.10
    0.5%
    738
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    399
    +0.18
    0.02
    275
    +0.10
    0.02
    257
    +0.10
    0.02
    Negative Logits
    aturdays
    -0.79
    ratulations
    -0.76
    tons
    -0.75
    merce
    -0.71
    FontSize
    -0.69
    pled
    -0.69
     deadlines
    -0.69
    inse
    -0.69
    AFTA
    -0.69
    success
    -0.67
    POSITIVE LOGITS
     whence
    0.87
     Flavoring
    0.75
     afar
    0.73
     Everett
    0.72
     Manga
    0.69
     Pastebin
    0.66
     Depths
    0.65
     EDIT
    0.64
     TheNitromeFan
    0.63
     Elk
    0.63
    Activations Density 0.021%

    No Known Activations