© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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. 18007
    Prev
    Next
    INDEX
    Explanations

    phrases related to social and political issues

    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
    481
    +0.26
    1.2%
    447
    +0.16
    0.7%
    19
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.26
    0.09
    481
    +0.16
    0.07
    19
    +0.12
    0.05
    Negative Logits
    thouse
    -0.80
    arnaev
    -0.74
    pired
    -0.68
     Airl
    -0.64
    bered
    -0.63
     Airbus
    -0.62
    lio
    -0.61
     Onion
    -0.60
    ERT
    -0.60
    rette
    -0.59
    POSITIVE LOGITS
     topics
    1.08
     etiquette
    0.98
     issues
    0.97
     matters
    0.96
     affordability
    0.89
     versus
    0.86
    Semitism
    0.86
     ethics
    0.85
     misconceptions
    0.85
     neutrality
    0.84
    Activations Density 0.583%

    No Known Activations