© 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. 104777
    Prev
    Next
    INDEX
    Explanations

    would/will followed by an action

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

    The neuron detects modal and future‐tense auxiliary verbs (e.g. will, would, ’ll, be, able, have) signaling modality or prospective action.

    oai_token-act-pair · o4-miniTriggered 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
    )$
    -1.52
     a
    -1.43
    :……
    -1.41
     */
    
    
    -1.38
     poate
    -1.36
    都会
    -1.34
    =[]
    
    -1.34
    ’).
    -1.33
    ↵
    -1.30
    FirebaseUser
    -1.30
    POSITIVE LOGITS
    tifies
    1.28
    ૃ
    1.26
     enigma
    1.25
    agers
    1.21
    hance
    1.20
     posibil
    1.20
    憋
    1.20
     ジップ
    1.19
     SPANISH
    1.17
    ूल
    1.16
    Activations Density 0.021%

    No Known Activations