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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 23916
    Prev
    Next
    INDEX
    Explanations

    references to the NFL and its associated players and games

    oai_token-act-pair · gpt-4o-miniTriggered 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
    481
    +0.25
    1.1%
    447
    +0.17
    0.8%
    11
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    11
    +0.25
    0.05
    288
    +0.17
    0.07
    652
    +0.11
    0.05
    Negative Logits
    antry
    -0.70
    lishes
    -0.69
    cule
    -0.66
    shapeshifter
    -0.65
    ainer
    -0.64
    ħĭ
    -0.60
    ahime
    -0.58
    aky
    -0.58
    izabeth
    -0.57
    flows
    -0.57
    POSITIVE LOGITS
     baseball
    1.06
     stadiums
    1.06
     soccer
    1.02
     basketball
    1.02
     football
    0.99
     stadium
    0.94
     hockey
    0.93
     jerseys
    0.90
     rugby
    0.89
     Football
    0.88
    Activations Density 0.309%

    No Known Activations