© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 3968
    Prev
    Next
    INDEX
    Explanations

    time-related information like specific dates and days of the week

    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%
    612
    +0.11
    0.5%
    481
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    430
    +0.15
    0.09
    306
    +0.11
    0.07
    115
    +0.10
    0.07
    Negative Logits
    GY
    -0.61
     destruction
    -0.58
     sew
    -0.58
     flour
    -0.58
    animate
    -0.58
     reuse
    -0.56
     sooner
    -0.55
    ISH
    -0.54
     EVERY
    -0.54
     penetration
    -0.54
    POSITIVE LOGITS
     outlining
    0.84
     announcing
    0.84
     afternoon
    0.82
     reassuring
    0.79
     stating
    0.78
     morning
    0.75
     evening
    0.74
     discussing
    0.74
    athing
    0.73
     saying
    0.73
    Activations Density 0.194%

    No Known Activations