© 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. 1292
    Prev
    Next
    INDEX
    Explanations

    mentions of specific organizations or entities related to government and international affairs

    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
    355
    +0.10
    0.4%
    314
    +0.09
    0.4%
    517
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.10
    0.06
    267
    +0.09
    0.05
    251
    +0.09
    0.04
    Negative Logits
    rencies
    -0.84
    ities
    -0.80
    paces
    -0.79
    ernels
    -0.79
    riages
    -0.78
    aters
    -0.77
    ularity
    -0.77
    ages
    -0.76
     Romans
    -0.76
    ads
    -0.76
    POSITIVE LOGITS
     staffer
    1.21
     spokeswoman
    1.20
     spokesman
    1.15
     spokesperson
    1.15
     colleague
    1.13
     employee
    1.04
     representative
    0.97
     broch
    0.96
     official
    0.95
     aide
    0.94
    Activations Density 0.222%

    No Known Activations