© 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. 552
    Prev
    Next
    INDEX
    Explanations

    sentences expressing gratitude, thanks, or appreciation

    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
    155
    +0.11
    0.5%
    415
    +0.11
    0.5%
    41
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    566
    +0.11
    0.05
    307
    +0.11
    0.04
    41
    +0.10
    0.04
    Negative Logits
     simultane
    -0.63
    Reports
    -0.63
     intangible
    -0.60
     accuser
    -0.59
     periodic
    -0.57
     evid
    -0.56
     nightly
    -0.56
    notation
    -0.54
     reminders
    -0.54
     DOC
    -0.54
    POSITIVE LOGITS
    schild
    0.90
     hesitated
    0.69
    vana
    0.67
    horn
    0.66
    croft
    0.66
     mattered
    0.65
     dreamed
    0.65
    heit
    0.63
    WARE
    0.63
    vation
    0.61
    Activations Density 0.161%

    No Known Activations