© 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. 4079
    Prev
    Next
    INDEX
    Explanations

    names of individuals, teams, or entities in the context of sports, competitions, or events

    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
    481
    +0.21
    1.0%
    447
    +0.20
    1.0%
    288
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.21
    0.07
    326
    +0.20
    0.05
    353
    +0.11
    0.04
    Negative Logits
    acters
    -0.80
    jriwal
    -0.76
    ounter
    -0.67
     charism
    -0.67
    perature
    -0.67
    otomy
    -0.66
     quo
    -0.65
    ulates
    -0.63
    nesses
    -0.62
     proble
    -0.62
    POSITIVE LOGITS
     Labs
    0.99
     Genetics
    0.83
     Technologies
    0.80
     Industries
    0.79
     Lounge
    0.78
    Hub
    0.78
     Analytics
    0.78
     Networks
    0.77
     Records
    0.77
     Airlines
    0.76
    Activations Density 0.257%

    No Known Activations