Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 26686
    Prev
    Next
    INDEX
    Explanations

    references to the television show "The Simpsons."

    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
    575
    +0.11
    0.5%
    699
    +0.10
    0.5%
    545
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    683
    +0.11
    0.01
    545
    +0.10
    0.01
    722
    +0.10
    0.01
    Negative Logits
    places
    -0.75
    inventoryQuantity
    -0.72
    arians
    -0.72
    arian
    -0.71
    abil
    -0.67
    enser
    -0.67
    inventory
    -0.66
    tips
    -0.66
    SOURCE
    -0.65
    IDA
    -0.64
    POSITIVE LOGITS
     Simpson
    1.12
    pson
    1.04
    psons
    0.78
    mie
    0.70
    inez
    0.68
     Falls
    0.68
     Clause
    0.67
     Broadcasting
    0.67
    ville
    0.65
     Bows
    0.64
    Activations Density 0.004%

    No Known Activations