© 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 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. 10304
    Prev
    Next
    INDEX
    Explanations

    phrases related to video game levels, particularly focusing on aspects like difficulty, design, and mechanics

    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
    99
    +0.12
    0.5%
    113
    +0.10
    0.5%
    678
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    99
    +0.12
    0.02
    198
    +0.10
    0.02
    647
    +0.10
    0.02
    Negative Logits
    vous
    -0.77
    ILLE
    -0.75
    fect
    -0.68
    uter
    -0.68
    orial
    -0.67
    orie
    -0.67
    ãĥĹ
    -0.64
     Leilan
    -0.64
    MRI
    -0.64
    oute
    -0.62
    POSITIVE LOGITS
     crossings
    0.81
     assurance
    0.75
     difference
    0.70
    level
    0.69
     attainment
    0.67
    rise
    0.66
     levels
    0.66
     attained
    0.65
     shif
    0.65
     distinctions
    0.65
    Activations Density 0.037%

    No Known Activations