© 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 49152
    9. 8-RES_FS49152-JB
    11. 45537
    Prev
    Next
    INDEX
    Explanations

    mentions of yoga and gymnastics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    288
    +0.15
    0.7%
    447
    +0.14
    0.6%
    343
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    343
    +0.15
    0.01
    288
    +0.14
    0.01
    57
    +0.12
    0.01
    Negative Logits
    hyde
    -0.82
    vous
    -0.73
    etary
    -0.73
    strous
    -0.72
    */(
    -0.71
    endez
    -0.71
    ictional
    -0.70
    ighty
    -0.69
    ilial
    -0.68
     Snowden
    -0.66
    POSITIVE LOGITS
     meditation
    0.99
     yoga
    0.98
     instructor
    0.96
     mats
    0.92
     pants
    0.89
    oga
    0.87
     instructors
    0.87
    nas
    0.86
     routines
    0.82
     Yoga
    0.82
    Activations Density 0.008%

    No Known Activations