© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 9177
    Prev
    Next
    INDEX
    Explanations

    can followed by a verb

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

    The neuron activates on modal auxiliary verbs expressing ability or possibility (e.g. “can,” “could,” “would”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    expressions of ability or possibility, especially modal constructions (including negatives) like “we/I can/could” and “can be.”

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    the modal verb "can" indicating ability or possibility.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    --
    
    -1.40
    k
    -1.32
    Эти
    -1.31
    édr
    -1.30
    茄
    -1.26
    瘤
    -1.26
     emballage
    -1.23
    ίας
    -1.23
    卓越
    -1.23
    дом
    -1.23
    POSITIVE LOGITS
    そういえば
    1.40
    お勧め
    1.33
    そうそう
    1.32
    出场
    1.32
    けっこう
    1.27
     able
    1.24
    坚定
    1.24
     possible
    1.23
     better
    1.19
    少々
    1.17
    Activations Density 0.083%

    No Known Activations