© 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. 5361
    Prev
    Next
    INDEX
    Explanations

    information related to historical events, specifically focusing on the evolution of industries and movements over time

    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
    679
    +0.15
    0.7%
    640
    +0.14
    0.7%
    447
    +0.14
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    6
    +0.15
    0.12
    679
    +0.14
    0.10
    440
    +0.14
    0.08
    Negative Logits
     slightest
    -0.75
    odo
    -0.72
     purse
    -0.71
     explan
    -0.71
     antiv
    -0.69
     shark
    -0.69
     zone
    -0.69
    bang
    -0.68
     cough
    -0.68
     selfie
    -0.68
    POSITIVE LOGITS
     Eventually
    1.62
     Later
    1.44
     During
    1.33
     Initially
    1.28
     Ultimately
    1.27
     Afterwards
    1.24
     Ironically
    1.24
     Throughout
    1.22
     Although
    1.22
    Eventually
    1.20
    Activations Density 0.406%

    No Known Activations