© 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. 37566
    Prev
    Next
    INDEX
    Explanations

    instances of the word "is."

    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
    195
    +0.14
    0.6%
    459
    +0.13
    0.6%
    578
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    304
    +0.14
    0.02
    659
    +0.13
    0.02
    700
    +0.12
    0.02
    Negative Logits
    ishers
    -0.69
    ¥ŀ
    -0.68
    itness
    -0.65
    ¥µ
    -0.65
    è¦ļéĨĴ
    -0.64
    ilight
    -0.63
    iage
    -0.62
    osponsors
    -0.61
    oat
    -0.61
     thumbnail
    -0.61
    POSITIVE LOGITS
    abeth
    1.31
    aurus
    1.05
    peed
    1.03
    earch
    0.99
    henko
    0.93
    terness
    0.92
    ection
    0.91
    ource
    0.91
    aur
    0.90
    ippi
    0.88
    Activations Density 0.026%

    No Known Activations