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 768
    5. 8-RES_FS768-JB
    6. 156
    Prev
    Next
    INDEX
    Explanations

    economic and financial terms and statistics

    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
    393
    +0.16
    0.8%
    484
    +0.11
    0.5%
    0
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    484
    +0.16
    0.18
    0
    +0.11
    0.18
    393
    +0.11
    0.20
    Negative Logits
    uta
    -0.88
    actionDate
    -0.75
    DX
    -0.73
    odcast
    -0.69
    ibble
    -0.69
    uct
    -0.69
    gest
    -0.68
    zip
    -0.68
    Street
    -0.68
    DragonMagazine
    -0.68
    POSITIVE LOGITS
     thus
    1.62
     consequently
    1.61
     hence
    1.58
     therefore
    1.53
     thereby
    1.49
     secondly
    1.41
     furthermore
    1.27
     thence
    1.23
     then
    1.19
     moreover
    1.16
    Activations Density 2.860%

    No Known Activations