© 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. 9595
    Prev
    Next
    INDEX
    Explanations

    the concept of ideas or propositions being discussed or expressed

    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
    339
    +0.14
    0.6%
    541
    +0.11
    0.5%
    641
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    247
    +0.14
    0.03
    472
    +0.11
    0.03
    641
    +0.10
    0.04
    Negative Logits
    administ
    -0.72
    Oper
    -0.68
    pex
    -0.65
     [+
    -0.64
    ×ŀ
    -0.64
    Naz
    -0.63
    ances
    -0.63
    aud
    -0.62
    attering
    -0.61
    ×IJ
    -0.61
    POSITIVE LOGITS
    ually
    0.96
     idea
    0.85
     hatched
    0.82
     Ideas
    0.78
     ideas
    0.78
     provoking
    0.78
     floated
    0.73
     concept
    0.72
     inspir
    0.71
     sugg
    0.71
    Activations Density 0.087%

    No Known Activations