© 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. 439
    Prev
    Next
    INDEX
    Explanations

    items or objects related to crafts or construction

    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
    721
    +0.12
    0.5%
    535
    +0.10
    0.4%
    60
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    60
    +0.12
    0.04
    306
    +0.10
    0.04
    551
    +0.10
    0.04
    Negative Logits
     striving
    -0.68
     Confeder
    -0.61
    é¾įå¥ij士
    -0.61
     rehears
    -0.60
     rehearsal
    -0.57
     reven
    -0.56
     aven
    -0.56
    eering
    -0.56
     revision
    -0.56
     honoring
    -0.55
    POSITIVE LOGITS
    't
    1.49
    berra
    1.41
    adian
    1.34
    vas
    1.29
    isters
    1.17
    opy
    1.11
    ister
    1.09
    ibus
    0.98
    nery
    0.98
    NOT
    0.97
    Activations Density 0.088%

    No Known Activations