© 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. 10201
    Prev
    Next
    INDEX
    Explanations

    information about show business and entertainment industry news

    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
    176
    +0.14
    0.7%
    479
    +0.12
    0.6%
    588
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    479
    +0.14
    0.04
    588
    +0.12
    0.06
    45
    +0.12
    0.04
    Negative Logits
    izabeth
    -0.92
    alysed
    -0.89
    iscal
    -0.83
    agonists
    -0.79
    earcher
    -0.73
    autions
    -0.72
    income
    -0.72
    istical
    -0.72
    duty
    -0.70
    agonist
    -0.70
    POSITIVE LOGITS
     Twisted
    1.01
     Magnum
    0.98
     Outbreak
    0.97
     Pet
    0.97
     Dinosaur
    0.96
     Chocolate
    0.95
     Glory
    0.95
     Classic
    0.95
     Lucky
    0.95
     Monkey
    0.95
    Activations Density 0.476%

    No Known Activations