© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 12352
    Prev
    Next
    INDEX
    Explanations

    fictional writing types

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

    The neuron fires on subjective evaluative language—especially strong opinion words and phrases (e.g. “love,” “excellent,” “recommend,” “noisy,” “little noisy”) that signal the reviewer’s positive or negative judgments.

    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
     delightfully
    0.54
     sumptuous
    0.53
      
    0.53
    The
    0.49
     The
    0.47
     hapless
    0.45
     thưởng
    0.44
     bulunmaktadır
    0.44
    傢
    0.44
    indul
    0.43
    POSITIVE LOGITS
     -"
    0.52
    ...'
    0.48
    ...')
    0.46
     BEACH
    0.44
     '--
    0.44
     Aguirre
    0.43
    '...
    0.43
    ]'
    0.43
     shogun
    0.43
    -'
    0.43
    Activations Density 0.075%

    No Known Activations