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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 517
    Prev
    Next
    INDEX
    Explanations

    mentions of specific places or events

    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
    483
    +0.11
    0.5%
    53
    +0.10
    0.4%
    626
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.11
    0.04
    232
    +0.10
    0.03
    373
    +0.09
    -0.01
    Negative Logits
     Matter
    -0.68
    BuyableInstoreAndOnline
    -0.67
     welcome
    -0.64
    20439
    -0.62
    AGES
    -0.59
     disadvant
    -0.58
     seiz
    -0.58
     Wonderland
    -0.57
    paces
    -0.57
     centrif
    -0.56
    POSITIVE LOGITS
    ciples
    1.15
    cipled
    1.07
    bably
    0.88
    fecture
    0.73
    issance
    0.72
    ety
    0.71
    cess
    0.71
    loo
    0.68
    ential
    0.68
    esses
    0.68
    Activations Density 0.128%

    No Known Activations