© 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. 10183
    Prev
    Next
    INDEX
    Explanations

    proper names, specifically focusing on the last names of individuals

    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
    430
    +0.09
    0.4%
    314
    +0.09
    0.4%
    180
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.09
    0.09
    659
    +0.09
    0.07
    667
    +0.09
    0.06
    Negative Logits
     COM
    -0.71
     indo
    -0.70
     operating
    -0.69
     prost
    -0.69
     metast
    -0.68
     supers
    -0.68
     TRA
    -0.68
     acron
    -0.67
     wast
    -0.67
     USS
    -0.67
    POSITIVE LOGITS
    ovich
    1.15
    ason
    1.09
    chuk
    1.06
    ault
    1.05
    wald
    1.04
    neau
    1.03
    iak
    1.03
    reau
    1.03
    bach
    1.03
    orst
    1.03
    Activations Density 0.226%

    No Known Activations