© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26
    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 3284
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on very short, high-frequency function words—most prominently the single-letter article “a” and the two-letter preposition “at.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    善
    0.63
    ഭ്യാ
    0.57
    橘
    0.56
     限
    0.52
     माइनस
    0.52
    ོ
    0.52
    Antaeotricha
    0.52
    鼻
    0.52
     Normalize
    0.51
     hopelessness
    0.51
    POSITIVE LOGITS
     increasingly
    0.64
     proportional
    0.57
     stadium
    0.57
    越来越
    0.57
     muze
    0.56
     accelerating
    0.55
     tetapi
    0.54
     faster
    0.53
     sages
    0.53
     dp
    0.52
    Activations Density 0.000%

    No Known Activations