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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlog/PodcastSlackPrivacy & 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. 136
    Prev
    Next
    INDEX
    Explanations

    information related to sports matches and player performances

    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
    281
    +0.13
    0.6%
    238
    +0.13
    0.6%
    603
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    281
    +0.13
    0.13
    238
    +0.13
    0.14
    603
    +0.12
    0.11
    Negative Logits
     memos
    -0.76
    ographies
    -0.76
     curric
    -0.76
    reprene
    -0.75
    aminer
    -0.74
     journals
    -0.74
    onym
    -0.74
    Companies
    -0.73
    minist
    -0.70
    nih
    -0.70
    POSITIVE LOGITS
     unbeaten
    1.45
     undefeated
    1.42
     playoff
    1.27
     rematch
    1.21
     halftime
    1.19
     playoffs
    1.14
     Playoff
    1.10
     relegation
    1.10
     championship
    1.08
     win
    1.07
    Activations Density 5.028%

    No Known Activations