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

    Neuronpedia

    APISteerSAE EvalsBlog/PodcastNEWSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlog/PodcastGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 8333
    Prev
    Next
    INDEX
    Explanations

    bookmarks toward the end of a text

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    438
    +0.13
    0.6%
    167
    +0.12
    0.5%
    651
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    651
    +0.13
    0.03
    438
    +0.12
    0.03
    354
    +0.11
    0.03
    Negative Logits
    ilitary
    -0.78
     VIDEOS
    -0.76
    asio
    -0.70
    00200000
    -0.64
    govtrack
    -0.63
     Santiago
    -0.61
    odder
    -0.60
    Downloadha
    -0.60
     Democr
    -0.60
     constitu
    -0.59
    POSITIVE LOGITS
    marks
    1.47
    stores
    1.43
    worm
    1.33
    seller
    1.27
    keeping
    1.27
    marked
    1.26
    cases
    1.23
    mark
    1.20
    shop
    1.20
    worms
    1.19
    Activations Density 0.037%

    No Known Activations