© 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. 1838
    Prev
    Next
    INDEX
    Explanations

    references related to documents, publishing, and articles

    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
    87
    +0.17
    0.8%
    393
    +0.15
    0.7%
    459
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.17
    0.15
    575
    +0.15
    0.12
    459
    +0.11
    0.10
    Negative Logits
    ertain
    -0.64
    ãĤ£
    -0.59
    sych
    -0.57
    Billy
    -0.56
    ãĥ£
    -0.56
     Refugee
    -0.54
    arn
    -0.53
    ildo
    -0.52
    agen
    -0.52
    osaurs
    -0.52
    POSITIVE LOGITS
     rul
    0.84
     confir
    0.76
     withd
    0.76
     destro
    0.69
     Azerb
    0.67
     horizont
    0.64
     Niet
    0.63
     tyr
    0.61
    etheless
    0.61
     proport
    0.61
    Activations Density 1.954%

    No Known Activations