Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 768
    5. 8-RES_FS768-JB
    6. 383
    Prev
    Next
    INDEX
    Explanations

    technical terms and descriptions related to web development or online content

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    480
    +0.70
    7.3%
    266
    +0.16
    1.6%
    481
    +0.12
    1.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.70
    0.09
    266
    +0.16
    0.43
    276
    +0.12
    0.15
    Negative Logits
     destro
    -0.37
    anwhile
    -0.34
    Sov
    -0.30
     Tanz
    -0.28
     Azerb
    -0.27
    retty
    -0.27
     Vaugh
    -0.27
    eday
    -0.25
     cryst
    -0.25
     awa
    -0.25
    POSITIVE LOGITS
     Belfast
    0.28
    cigarettes
    0.23
     frontman
    0.23
     Donald
    0.22
     Hillary
    0.22
     Editor
    0.22
     Kavanaugh
    0.22
    hillary
    0.21
     Theresa
    0.21
    kowski
    0.21
    Activations Density 21.215%

    No Known Activations