© 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. 13179
    Prev
    Next
    INDEX
    Explanations

    search queries and structured data

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

    The neuron is tuned to pick up on second‐person hypothetical/conditional phrasing—i.e. tokens like “if,” “you,” modals (“could,” “might,” “have”), etc., that occur in “if you ever…” style constructions.

    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_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
    
    0.77
    etzte
    0.71
    éon
    0.70
    ریان
    0.69
    зя
    0.66
    alini
    0.65
     নিষ
    0.64
     ʼ
    0.63
     chaotic
    0.62
    ców
    0.62
    POSITIVE LOGITS
     Asthma
    0.90
     atrophy
    0.89
     Quito
    0.88
     eczema
    0.87
     Notify
    0.87
     keduanya
    0.85
     reaksi
    0.84
     Conference
    0.82
     anemia
    0.82
     suami
    0.81
    Activations Density 0.001%

    No Known Activations