© 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. 10362
    Prev
    Next
    INDEX
    Explanations

    past tense verbs indicating actions performed by specific 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
    566
    +0.15
    0.7%
    102
    +0.12
    0.6%
    377
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    566
    +0.15
    0.08
    55
    +0.12
    0.08
    102
    +0.12
    0.07
    Negative Logits
     Highlands
    -0.65
    acters
    -0.64
    ertodd
    -0.62
     lur
    -0.61
     Highlights
    -0.61
    =-=-
    -0.60
     Valhalla
    -0.58
     largeDownload
    -0.55
     override
    -0.55
    utenant
    -0.55
    POSITIVE LOGITS
     supposed
    1.26
     doing
    1.12
     talking
    1.02
     accustomed
    1.01
     trying
    1.00
     going
    1.00
     aiming
    0.99
     gonna
    0.98
     able
    0.98
     experiencing
    0.98
    Activations Density 0.218%

    No Known Activations