© 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
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 11039
    Prev
    Next
    INDEX
    Explanations

    gear you use

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

    This neuron detects abstract, nominalized content words—i.e. nouns formed from actions or concepts (like “commentary,” “retweet,” “footage,” “coverage,” “research,” “implications,” “technology,” etc.).

    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_31_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
    ನಿ
    0.66
    宪
    0.64
    琦
    0.63
    weapon
    0.62
    retval
    0.62
    سٹم
    0.61
    䢙
    0.61
    வி
    0.59
    寬
    0.58
    权力
    0.57
    POSITIVE LOGITS
     pieces
    0.82
     chunks
    0.74
     levels
    0.69
     Chunk
    0.66
     crowns
    0.65
    ൊന്നും
    0.65
    pieces
    0.65
     patterns
    0.64
     চাহিল
    0.64
     situations
    0.62
    Activations Density 0.557%

    No Known Activations