© 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. 298
    Prev
    Next
    INDEX
    Explanations

    programming-related terms or actions

    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
    345
    +0.09
    0.4%
    197
    +0.09
    0.4%
    728
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    665
    +0.09
    0.05
    764
    +0.09
    0.03
    641
    +0.08
    0.04
    Negative Logits
    dt
    -0.71
    mobi
    -0.70
    none
    -0.69
    von
    -0.68
    esm
    -0.67
    letter
    -0.67
    elin
    -0.67
    txt
    -0.67
    html
    -0.65
    aires
    -0.64
    POSITIVE LOGITS
     assignments
    0.78
    tains
    0.69
     oneself
    0.65
     strangers
    0.64
    RAG
    0.63
     these
    0.63
     transitions
    0.63
     Directions
    0.63
     certain
    0.61
     unfamiliar
    0.61
    Activations Density 0.320%

    No Known Activations