© 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. 9587
    Prev
    Next
    INDEX
    Explanations

    websites and online sources

    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
    766
    +0.12
    0.5%
    107
    +0.11
    0.5%
    496
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    586
    +0.12
    0.05
    107
    +0.11
    0.04
    766
    +0.11
    0.04
    Negative Logits
     Conclusion
    -0.83
     Goldberg
    -0.76
     Pants
    -0.75
     Advertisement
    -0.74
     Pie
    -0.73
     Caption
    -0.73
     Stretch
    -0.73
     Gamble
    -0.72
     Correction
    -0.71
     Singer
    -0.71
    POSITIVE LOGITS
    cdn
    1.26
    online
    1.14
    archives
    1.08
    nexus
    1.04
    foundation
    1.04
    research
    1.00
    archive
    0.99
    library
    0.96
    network
    0.95
    pedia
    0.95
    Activations Density 0.140%

    No Known Activations