© 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. 9721
    Prev
    Next
    INDEX
    Explanations

    videos or mentions of videos

    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
    447
    +0.21
    1.0%
    346
    +0.10
    0.5%
    690
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    644
    +0.21
    0.04
    346
    +0.10
    0.04
    690
    +0.09
    0.04
    Negative Logits
    merce
    -0.99
    cffff
    -0.98
    iders
    -0.79
    forces
    -0.77
    utenberg
    -0.76
    elsius
    -0.73
    Ĭ±
    -0.69
    arta
    -0.69
     accommodation
    -0.68
    athered
    -0.68
    POSITIVE LOGITS
     clip
    1.25
     footage
    1.06
     uploaded
    1.04
     filmed
    1.02
     clips
    1.02
     mont
    1.01
     depicting
    1.00
     snippet
    0.95
     aired
    0.95
     screenshot
    0.92
    Activations Density 0.168%

    No Known Activations