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

    Neuronpedia

    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 12288
    9. 8-RES_FS12288-JB
    11. 9454
    Prev
    Next
    INDEX
    Explanations

    phrases related to politics and government

    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
    12
    +0.11
    0.5%
    665
    +0.11
    0.5%
    291
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    8
    +0.11
    0.06
    598
    +0.11
    0.04
    233
    +0.10
    0.04
    Negative Logits
    cellaneous
    -0.70
    --+
    -0.66
    Gene
    -0.64
    isable
    -0.63
    ieth
    -0.62
    unts
    -0.61
    ileged
    -0.60
    lear
    -0.60
    venge
    -0.60
    ablishment
    -0.59
    POSITIVE LOGITS
     actually
    1.08
     truth
    1.03
    Actually
    1.01
    actually
    0.98
     reality
    0.98
     Actually
    0.97
     Fraud
    0.85
     actual
    0.84
    agine
    0.77
     facts
    0.76
    Activations Density 0.493%

    No Known Activations