© 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
    2. Gemma-3-27B-IT
    3. 31-GEMMASCOPE-2-RES-262K
    4. 32255
    Prev
    Next
    INDEX
    Explanations

    sarcastic or ironically positive

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

    sarcastic, snarky conversational asides and emphatic interjections that signal irony or attitude.

    oai_token-act-pair · gpt-5Triggered by @vetterc0

    Sarcastic or ironic phrases, often marked by specific words like "Oh," "wonderful," "fantastic," "fun," or "interesting," typically used to express the opposite of their literal meaning, frequently appearing in contexts involving feigned enthusiasm, mock politeness, or humorous dismissal of requests or situations.

    eleuther_acts_top20 · claude-4-5-sonnetTriggered by @jamesnaruto04
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/resid_post/layer_31_width_262k_l0_medium
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     metaphorical
    0.42
     obs
    0.39
     theta
    0.37
    次元
    0.37
    IPS
    0.37
     emergence
    0.37
     steroids
    0.37
    اتج
    0.37
     span
    0.36
     metaphors
    0.36
    POSITIVE LOGITS
     infallible
    0.52
    perfect
    0.50
     perfetta
    0.46
     innocent
    0.46
    🤡
    0.46
     foolproof
    0.45
     sempurna
    0.45
     perfettamente
    0.45
     perfeita
    0.44
    하겠습니다
    0.43
    Activations Density 0.075%

    No Known Activations