© 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
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 10978
    Prev
    Next
    INDEX
    Explanations

    reporting speech

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

    This neuron detects reporter‐style attribution phrases, i.e. occurrences of a person’s name (or pronoun) followed by “said.”

    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
     iż
    1.19
    یان
    1.15
    იც
    1.14
     البته
    1.13
    DLJ
    1.11
    Ду
    1.09
    𝐂
    1.09
    Uttar
    1.07
    Чи
    1.05
     faptul
    1.05
    POSITIVE LOGITS
    tte
    1.10
    led
    1.06
    ed
    1.05
    tion
    1.02
    )
    1.01
    $
    1.00
    ly
    0.96
     known
    0.94
    _
    0.94
    ன்
    0.93
    Activations Density 0.017%

    No Known Activations