© 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. 468
    Prev
    Next
    INDEX
    Explanations

    references to outcomes or conclusions

    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
    83
    +0.12
    0.5%
    129
    +0.10
    0.5%
    679
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    540
    +0.12
    0.04
    542
    +0.10
    0.03
    83
    +0.10
    0.03
    Negative Logits
     Passage
    -0.70
    conservancy
    -0.69
    afort
    -0.68
     actionGroup
    -0.68
    resent
    -0.65
    erness
    -0.63
    agine
    -0.62
     Roots
    -0.61
    kov
    -0.61
    nt
    -0.60
    POSITIVE LOGITS
    iveness
    0.94
    ivity
    0.87
    ively
    0.85
    aries
    0.80
    results
    0.79
     results
    0.79
     thereof
    0.76
    iments
    0.75
     result
    0.73
    ainer
    0.70
    Activations Density 0.073%

    No Known Activations