© 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 24576
    9. 8-RES_FS24576-JB
    11. 20066
    Prev
    Next
    INDEX
    Explanations

    proper nouns, specifically names of politicians and government officials

    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
    224
    +0.14
    0.6%
    31
    +0.11
    0.5%
    700
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.14
    0.06
    224
    +0.11
    0.06
    700
    +0.10
    0.05
    Negative Logits
     scouts
    -0.76
    Reviewer
    -0.72
    olars
    -0.71
    books
    -0.69
     experien
    -0.68
     civilisation
    -0.67
     skelet
    -0.67
    translation
    -0.67
    properties
    -0.67
    tainment
    -0.66
    POSITIVE LOGITS
    Senate
    1.08
     Jr
    1.01
     Caucus
    1.00
     filib
    0.95
    ospons
    0.92
     filibuster
    0.92
     voted
    0.91
    Republican
    0.91
     reelection
    0.90
    govtrack
    0.90
    Activations Density 0.153%

    No Known Activations