© 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. 37526
    Prev
    Next
    INDEX
    Explanations

    references to the concept of a "sham" or fraudulent situations

    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
    +0.16
    0.8%
    291
    +0.11
    0.5%
    50
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    444
    +0.16
    0.01
    291
    +0.11
    0.01
    288
    +0.11
    0.01
    Negative Logits
    inventoryQuantity
    -0.68
     Crunch
    -0.64
     Extrem
    -0.62
    enhagen
    -0.59
     Garn
    -0.58
     Motorsport
    -0.58
     AVG
    -0.58
     explos
    -0.57
     EW
    -0.57
     Reach
    -0.57
    POSITIVE LOGITS
    blers
    1.11
    rock
    1.03
    ument
    1.01
    eless
    0.94
    ble
    0.92
    arak
    0.90
    urai
    0.90
    isen
    0.88
    aji
    0.83
    urn
    0.83
    Activations Density 0.007%

    No Known Activations