© 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 12288
    9. 8-RES_FS12288-JB
    11. 1753
    Prev
    Next
    INDEX
    Explanations

    phrases related to opposition or conflicting situations

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    496
    +0.12
    0.6%
    608
    +0.12
    0.5%
    172
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.12
    0.04
    34
    +0.12
    0.03
    281
    +0.10
    0.03
    Negative Logits
    ulously
    -0.83
    zl
    -0.83
    hops
    -0.83
    ulous
    -0.81
    ounce
    -0.79
    oola
    -0.79
    chell
    -0.78
    chin
    -0.75
    gow
    -0.73
    opher
    -0.73
    POSITIVE LOGITS
     whom
    0.96
     adversity
    0.84
     backdrop
    0.84
     Humanity
    0.83
     anybody
    0.80
     anyone
    0.72
     them
    0.71
     behalf
    0.69
     prejudice
    0.69
     humanity
    0.68
    Activations Density 0.082%

    No Known Activations