© 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. 4977
    Prev
    Next
    INDEX
    Explanations

    terms related to medical research and treatments

    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
    659
    +0.13
    0.6%
    167
    +0.11
    0.5%
    561
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.13
    0.03
    588
    +0.11
    0.03
    746
    +0.11
    0.02
    Negative Logits
     lapse
    -0.74
     CPC
    -0.70
     remembrance
    -0.68
    etheless
    -0.67
    terday
    -0.66
     wink
    -0.65
     disarm
    -0.62
    swer
    -0.62
     standby
    -0.62
     conditioning
    -0.61
    POSITIVE LOGITS
    ulpt
    1.36
    ratch
    1.35
    apers
    1.28
    attered
    1.22
    reens
    1.21
    andal
    1.20
    enario
    1.20
    ouring
    1.15
    rupulous
    1.14
    atters
    1.12
    Activations Density 0.012%

    No Known Activations