© 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. 23961
    Prev
    Next
    INDEX
    Explanations

    references to errors or issues encountered in processes or data

    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
    263
    +0.09
    0.4%
    502
    +0.09
    0.4%
    388
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    655
    +0.09
    0.02
    156
    +0.09
    0.01
    255
    +0.09
    0.01
    Negative Logits
    apy
    -0.82
    tsky
    -0.81
    electric
    -0.80
    nai
    -0.80
    ILA
    -0.77
    Electric
    -0.75
    edom
    -0.72
    amen
    -0.72
    apeake
    -0.71
    atos
    -0.70
    POSITIVE LOGITS
    ously
    0.95
    uracy
    0.92
     margin
    0.88
    fully
    0.81
    gered
    0.81
     prone
    0.79
    hooting
    0.79
     mishand
    0.78
    prone
    0.77
     corrected
    0.75
    Activations Density 0.030%

    No Known Activations