© 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 49152
    9. 8-RES_FS49152-JB
    11. 45637
    Prev
    Next
    INDEX
    Explanations

    occurrences of strong emotional expressions or emphatic phrases

    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
    314
    +0.16
    0.8%
    68
    +0.11
    0.5%
    373
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.16
    0.03
    621
    +0.11
    0.03
    68
    +0.09
    0.02
    Negative Logits
     tremend
    -0.74
     Downs
    -0.71
     assemb
    -0.70
     whistle
    -0.69
     decomp
    -0.67
     dispers
    -0.65
     dirt
    -0.65
     bearer
    -0.64
     promul
    -0.64
     straw
    -0.64
    POSITIVE LOGITS
    į
    1.07
    Ķ
    1.02
    ı
    1.01
    ¤
    1.00
    ł
    1.00
    ¬
    0.99
    º
    0.97
    ĸ
    0.96
    ľ
    0.95
    Ĥ
    0.94
    Activations Density 0.045%

    No Known Activations