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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 43341
    Prev
    Next
    INDEX
    Explanations

    legislative actions related to passing bills

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    665
    +0.11
    0.5%
    339
    +0.11
    0.5%
    615
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    761
    +0.11
    0.02
    350
    +0.11
    0.02
    665
    +0.10
    0.02
    Negative Logits
     Merrill
    -0.72
     awaken
    -0.64
    antha
    -0.63
     Brist
    -0.63
     Pie
    -0.62
    plug
    -0.62
     awoken
    -0.61
     istg
    -0.58
    Tur
    -0.58
     Forsaken
    -0.57
    POSITIVE LOGITS
     muster
    1.15
     legislation
    0.95
     unanimously
    0.94
    smanship
    0.90
     Legislation
    0.88
    interstitial
    0.85
    engers
    0.81
     resolutions
    0.81
    wd
    0.79
    ivity
    0.78
    Activations Density 0.017%

    No Known Activations