© 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. 10432
    Prev
    Next
    INDEX
    Explanations

    references to the former US president, Barack Obama

    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
    200
    +0.11
    0.5%
    15
    +0.10
    0.4%
    254
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    200
    +0.11
    0.05
    346
    +0.10
    0.04
    15
    +0.10
    0.04
    Negative Logits
     Cth
    -0.79
    asma
    -0.76
    itionally
    -0.76
     ILCS
    -0.75
    auri
    -0.74
    TOR
    -0.73
     Edinburgh
    -0.72
    âĵĺ
    -0.70
     Newcastle
    -0.69
    ategory
    -0.69
    POSITIVE LOGITS
    Care
    0.98
     administration
    0.93
     appoint
    0.88
    Obama
    0.86
     Administration
    0.86
    omics
    0.82
     Obama
    0.81
     vetoed
    0.80
    mire
    0.78
    's
    0.75
    Activations Density 0.082%

    No Known Activations