© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 12288
    9. 8-RES_FS12288-JB
    11. 10704
    Prev
    Next
    INDEX
    Explanations

    detailed technical instructions and software update information

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    640
    +0.18
    0.9%
    430
    +0.18
    0.8%
    6
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    6
    +0.18
    0.13
    430
    +0.18
    0.12
    640
    +0.13
    0.11
    Negative Logits
     imperson
    -0.79
    roph
    -0.77
     grit
    -0.77
     butcher
    -0.76
     boarding
    -0.75
     fingert
    -0.74
     imitation
    -0.73
     disguise
    -0.73
     harassing
    -0.72
     kindness
    -0.72
    POSITIVE LOGITS
     Previously
    1.17
     Afterwards
    1.14
     Dates
    1.10
     During
    1.09
     Congratulations
    1.08
     Immediately
    1.08
     Assuming
    1.06
     However
    1.06
     Shortly
    1.05
     Prior
    1.05
    Activations Density 0.573%

    No Known Activations