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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 10729
    Prev
    Next
    INDEX
    Explanations

    see you

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

    This neuron fires on occurrences of the verb “see” (and its past‐tense variant “saw”), especially when it appears as a standalone word or title.

    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
     incipient
    0.54
     Preparation
    0.54
    Computing
    0.52
    ('-');
    0.51
    вата
    0.51
     Addressing
    0.51
    vinced
    0.51
     Computation
    0.50
    Computed
    0.49
    叭
    0.49
    POSITIVE LOGITS
     Heterocycl
    0.75
     Thứ
    0.63
    ymptoms
    0.63
     واپسی
    0.63
     आजीवन
    0.62
     compétition
    0.61
     tjän
    0.61
     exteriores
    0.60
     günler
    0.60
    峼
    0.59
    Activations Density 0.001%

    No Known Activations