© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 16793
    Prev
    Next
    INDEX
    Explanations

    mentions of sports teams, games, and player performance

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    279
    +0.11
    0.5%
    650
    +0.08
    0.4%
    654
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    279
    +0.11
    0.08
    526
    +0.08
    -0.01
    762
    +0.08
    0.07
    Negative Logits
    Flickr
    -0.95
     NYT
    -0.88
    Requires
    -0.80
    âĢł
    -0.78
    NOW
    -0.78
    utterstock
    -0.76
     Salon
    -0.76
     ®
    -0.74
     Economist
    -0.74
     Lovecraft
    -0.74
    POSITIVE LOGITS
     coaching
    1.22
     adversity
    1.15
     teammates
    1.14
     coaches
    1.13
     offensively
    1.09
    Coach
    1.07
     Coach
    1.07
     preseason
    1.06
     coach
    1.05
     competitiveness
    1.04
    Activations Density 0.848%

    No Known Activations