© 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. 16839
    Prev
    Next
    INDEX
    Explanations

    dates represented in ordinal or cardinal form

    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
    430
    +0.15
    0.7%
    310
    +0.14
    0.6%
    306
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    310
    +0.15
    0.10
    306
    +0.14
    0.08
    430
    +0.11
    0.07
    Negative Logits
     explan
    -0.57
     hobbies
    -0.57
     tut
    -0.56
     streng
    -0.55
     habitual
    -0.54
     Arche
    -0.54
     sys
    -0.53
     lawy
    -0.53
     departments
    -0.53
     Genie
    -0.53
    POSITIVE LOGITS
    th
    1.54
    rd
    1.14
    ths
    1.05
    TH
    0.96
    teenth
    0.90
    nd
    0.89
    tha
    0.88
    ember
    0.88
    eteenth
    0.86
    st
    0.83
    Activations Density 0.102%

    No Known Activations