© 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. 19564
    Prev
    Next
    INDEX
    Explanations

    phrases related to sports injuries and players being unavailable due to physical issues

    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
    570
    +0.10
    0.4%
    511
    +0.09
    0.4%
    681
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    410
    +0.10
    0.02
    504
    +0.09
    0.02
    651
    +0.09
    0.02
    Negative Logits
    Reviewer
    -0.78
    Male
    -0.72
    eli
    -0.70
    arist
    -0.68
    oplan
    -0.67
    truth
    -0.66
    Els
    -0.64
     behaved
    -0.63
    天
    -0.62
    pa
    -0.61
    POSITIVE LOGITS
     curfew
    0.88
     indefinitely
    0.81
     surgery
    0.78
     awhile
    0.78
     prolonged
    0.73
     precaution
    0.73
     playoffs
    0.72
     extended
    0.72
     dearly
    0.71
     Surgery
    0.70
    Activations Density 0.036%

    No Known Activations