© 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. 2065
    Prev
    Next
    INDEX
    Explanations

    informal discourse markers

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

    This neuron responds to personal‐perspective markers—especially first‐ and second‐person pronouns and deictic tokens used in narration or spoken/instructional passages.

    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_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
    さまざまな
    1.15
    Asimismo
    1.05
     Messrs
    1.05
     variously
    1.05
    此外
    1.02
     نیز
    1.00
     впоследствии
    1.00
     wówczas
    1.00
     および
    1.00
    および
    1.00
    POSITIVE LOGITS
     crappy
    1.63
     alot
    1.48
     shitty
    1.48
    ってる
    1.44
     pissed
    1.41
     gotta
    1.40
     kinda
    1.38
    썻
    1.38
    してる
    1.37
     xbox
    1.34
    Activations Density 1.902%

    No Known Activations