© 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. 22-GEMMASCOPE-RES-131K
    4. 70960
    Prev
    Next
    INDEX
    Explanations

    no negation or directness

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

    The neuron strongly activates on the key noun or verb that immediately follows a negation (e.g. “no time left,” “no equivocation,” “don’t sugar-coat,” “no hidden charges,” “no excuses”), i.e. the word that names what is being denied, refused, or absent.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     pantaloni
    -1.19
    きっかけ
    -1.13
     vẫn
    -1.13
    삭제
    -1.11
     something
    -1.10
    로그인
    -1.09
     rather
    -1.07
    でもあります
    -1.05
    rather
    -0.98
    Ejemplo
    -0.98
    POSITIVE LOGITS
     or
    1.88
     anymore
    1.43
     this
    1.23
     here
    1.23
     nor
    1.19
     any
    1.16
     second
    1.05
     или
    1.03
    there
    0.98
     только
    0.96
    Activations Density 0.022%

    No Known Activations