© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 6600
    Prev
    Next
    INDEX
    Explanations

    vary by, not the

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron activates on terminology that expresses quantitative relationships or magnitudes (e.g. “proportional,” “increases,” “grows,” “negligible”).

    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
    FindingsRequest
    0.73
     jederzeit
    0.68
     друзьями
    0.65
    โน
    0.64
     полноцен
    0.64
     Citizens
    0.63
     utilizzando
    0.63
    ̤
    0.63
     ciudadanos
    0.63
     Mujeres
    0.63
    POSITIVE LOGITS
     proportional
    1.70
     proportionality
    1.54
     proporcional
    1.44
    ∝
    1.38
    propto
    1.33
     proportionate
    1.27
     пропор
    1.23
    portional
    1.20
     proportionally
    1.17
     exponential
    1.13
    Activations Density 0.151%

    No Known Activations