© 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. 5764
    Prev
    Next
    INDEX
    Explanations

    many or very

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

    The neuron fires on common perception or cognition verbs (e.g. know, see, get, feel, look) used to express subjective stance or opinion.

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

    negative contractions, particularly "didn't" and "don't".

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

    statements with grammatical negation, especially negated verb phrases expressing not doing/knowing/having/seeing or similar actions.

    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_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
     almost
    0.83
    almost
    0.80
     not
    0.72
     both
    0.70
     presque
    0.70
     sogar
    0.69
     cả
    0.68
    事实上
    0.68
    Almost
    0.67
     actually
    0.66
    POSITIVE LOGITS
     كثير
    1.02
     sondern
    0.99
     muitas
    0.98
     muchas
    0.96
     anymore
    0.93
     muchos
    0.88
     många
    0.88
     كثيرا
    0.88
    many
    0.87
     muitos
    0.87
    Activations Density 0.465%

    No Known Activations