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. 432
    Prev
    Next
    INDEX
    Explanations

    email addresses and social media usernames

    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
    481
    +0.11
    0.5%
    68
    +0.11
    0.5%
    23
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.11
    0.21
    621
    +0.11
    0.18
    317
    +0.10
    0.18
    Negative Logits
     Mous
    -0.87
     shroud
    -0.76
     Franch
    -0.71
     ASC
    -0.71
     Manit
    -0.70
     dispers
    -0.69
     cottage
    -0.67
     promul
    -0.67
     shack
    -0.65
     scatter
    -0.65
    POSITIVE LOGITS
    ľ
    1.67
    Ŀ
    1.60
    ¡
    1.56
    ª
    1.38
    ĺ
    1.33
    Ķ
    1.32
    ´
    1.31
    ¤
    1.30
    ¦
    1.29
    Ń
    1.28
    Activations Density 1.325%

    No Known Activations