© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 14801
    Prev
    Next
    INDEX
    Explanations

    are helpful

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

    The neuron strongly responds to descriptive predicate adjectives or past-participle modifiers—especially those following a copular verb (e.g. “are challenging,” “is super easy,” “were designed”).

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

    the word "are" when it appears before adjectives that describe positive qualities or characteristics.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    be-verb constructions functioning as copulas or passive auxiliaries, linking subjects to predicate adjectives or past participles.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_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
     해서
    0.81
    Então
    0.76
     पनि
    0.73
     участка
    0.73
    Há
    0.72
    penalty
    0.71
    Ná
    0.71
    блі
    0.71
    elemento
    0.71
    Mientras
    0.71
    POSITIVE LOGITS
     incorporates
    0.91
     accompanies
    0.89
     includes
    0.87
     resembles
    0.81
     contains
    0.75
     copied
    0.75
     accompany
    0.74
     shares
    0.73
     participated
    0.73
     depicts
    0.73
    Activations Density 0.887%

    No Known Activations