© 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. 24-GEMMASCOPE-2-RES-16K
    7. 1340
    Prev
    Next
    INDEX
    Explanations

    great / excellent / best

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

    The neuron responds strongly to positive evaluative adjectives (e.g. “great,” “amazing,” “phenomenal,” “outstanding”).

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

    the word "great" when used as a positive adjective to describe people, things, or qualities.

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

    words that express strong praise or high quality, especially emphatic evaluative adjectives and superlatives.

    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
    וש
    0.57
    บ่ง
    0.56
    窣
    0.55
    ﻒ
    0.55
    fang
    0.54
     불구하고
    0.53
     ഉപയോഗ
    0.53
     целей
    0.53
     Lemma
    0.52
     `:`
    0.52
    POSITIVE LOGITS
     meravigli
    0.83
     hebat
    0.73
     magnifique
    0.72
    wonderful
    0.71
    YOU
    0.71
     великолеп
    0.68
    无比
    0.67
     அற்புதமான
    0.67
    Starts
    0.66
     kvinn
    0.66
    Activations Density 0.002%

    No Known Activations