© 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 EvalsExportsinterp-engineNEWAPI 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. 4543
    Prev
    Next
    INDEX
    Explanations

    descriptions of products and services, particularly related to gaming, technology, and food

    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
    373
    +0.20
    0.9%
    447
    +0.15
    0.7%
    450
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    594
    +0.20
    0.04
    326
    +0.15
    0.05
    548
    +0.11
    0.02
    Negative Logits
    akable
    -0.72
    ofi
    -0.65
    gage
    -0.62
    Bey
    -0.62
    olulu
    -0.61
    poke
    -0.61
     thereof
    -0.60
    atern
    -0.59
     pony
    -0.58
    oday
    -0.58
    POSITIVE LOGITS
     Basics
    1.34
     Overview
    1.32
     Facts
    1.20
     Timeline
    1.18
     Comparison
    1.12
     Uses
    1.12
     Usage
    1.10
     Conclusion
    1.10
     Specifications
    1.08
     Versus
    1.06
    Activations Density 0.364%

    No Known Activations