© 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. 8044
    Prev
    Next
    INDEX
    Explanations

    text related to criticizing or mocking others

    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
    480
    +0.43
    2.2%
    266
    +0.10
    0.5%
    282
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.43
    0.01
    266
    +0.10
    0.08
    351
    +0.09
    0.04
    Negative Logits
     Warehouse
    -0.72
     rebuilt
    -0.71
     Located
    -0.67
    romeda
    -0.64
     pioneering
    -0.64
    chnology
    -0.61
    bitious
    -0.61
    phalt
    -0.61
    ufact
    -0.60
    erenn
    -0.60
    POSITIVE LOGITS
     sarcastic
    1.02
     slurs
    1.00
     ridicule
    0.98
     misinterpret
    0.96
     misunderstand
    0.95
     insults
    0.92
     insulting
    0.92
     replies
    0.91
     jokes
    0.90
     condesc
    0.89
    Activations Density 0.862%

    No Known Activations