© 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. 9890
    Prev
    Next
    INDEX
    Explanations

    good/great/perfect for

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

    The neuron fires strongly on emphatic positive praise—especially superlative adjectives like “perfect” (and similar words such as “brilliant,” “ideal,” or “most”) used to endorse or rave about something.

    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_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
    en
    1.01
     durchaus
    0.91
    7
    0.88
    2
    0.88
    ر
    0.85
    9
    0.84
    6
    0.84
    ن
    0.84
    www
    0.81
    enje
    0.80
    POSITIVE LOGITS
    soluble
    1.03
    ремя
    0.92
    თვის
    0.89
     idéal
    0.89
    흩
    0.87
     signalé
    0.86
     antidote
    0.86
     للع
    0.85
     opportune
    0.85
     ਲਈ
    0.84
    Activations Density 0.074%

    No Known Activations