© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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 49152
    9. 8-RES_FS49152-JB
    11. 42895
    Prev
    Next
    INDEX
    Explanations

    references to independence or autonomy in political contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    288
    +0.17
    0.8%
    263
    +0.12
    0.5%
    673
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    263
    +0.17
    0.02
    288
    +0.12
    0.01
    673
    +0.12
    0.01
    Negative Logits
    bilt
    -0.85
    ergy
    -0.79
    enegger
    -0.78
    geries
    -0.74
    ãĥ¥
    -0.73
    geon
    -0.72
    mers
    -0.69
    Commerce
    -0.68
    nos
    -0.67
    tail
    -0.67
    POSITIVE LOGITS
     referendum
    1.10
     oath
    0.85
     independence
    0.81
     Catalonia
    0.79
    yip
    0.74
     republic
    0.71
    inally
    0.69
     declaration
    0.68
     candidacy
    0.68
    heid
    0.68
    Activations Density 0.012%

    No Known Activations