© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 668
    Prev
    Next
    INDEX
    Explanations

    This neuron detects words ending in the suffix “-al,” i.e. adjectival forms like financial, ethical, physical, emotional, cultural, etc.

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

    words related to abstract concepts or intangible domains like finance, ethics, spirituality, culture, emotions, philosophy, and art.

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

    adjectives denoting abstract domains or qualities, often with -al/-ic/-ical suffixes.

    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
    oiden
    0.49
     Lover
    0.49
     Depuis
    0.47
     Başkanı
    0.46
    artige
    0.46
    ńca
    0.45
     की
    0.45
     Phần
    0.45
     allgemeinen
    0.45
    ansk
    0.44
    POSITIVE LOGITS
     prowess
    0.63
     aspects
    0.61
     barriers
    0.61
     upheaval
    0.59
    ரீ
    0.58
     fortitude
    0.58
     integrity
    0.57
     soundness
    0.57
    ity
    0.56
     correctness
    0.55
    Activations Density 0.598%

    No Known Activations