© 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. 1443
    Prev
    Next
    INDEX
    Explanations

    phrases related to debate and discussion

    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
    288
    +0.16
    0.7%
    447
    +0.13
    0.6%
    519
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    519
    +0.16
    0.03
    288
    +0.13
    0.04
    337
    +0.12
    0.03
    Negative Logits
    oho
    -0.69
    fits
    -0.67
    uras
    -0.65
    ridges
    -0.64
    imus
    -0.64
    photos
    -0.61
    iaries
    -0.61
    metics
    -0.60
    urances
    -0.60
    ionage
    -0.60
    POSITIVE LOGITS
     moderators
    0.92
     raged
    0.89
    halla
    0.88
     raging
    0.88
     debates
    0.86
     debate
    0.85
     forum
    0.83
     debating
    0.81
     concerning
    0.81
     moder
    0.81
    Activations Density 0.112%

    No Known Activations