© 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-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 84274
    Prev
    Next
    INDEX
    Explanations

    RM

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

    The neuron strongly activates on the uppercase two‐letter abbreviation “RM” (and the digits immediately following it), i.e. code or measure identifiers like “RM308,” “RM302,” etc.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    }
    -3.22
    肫
    -2.98
    ).
    -2.81
    骘
    -2.64
    kemeja
    -2.59
    üşonlu
    -2.47
     装修
    -2.47
    '].
    -2.45
    ⵔ
    -2.44
    ]
    -2.38
    POSITIVE LOGITS
    莛
    2.48
    曾在
    2.42
    发言
    2.36
     he
    2.33
     一个
    2.25
    莇
    2.23
     وفي
    2.23
    国庆
    2.19
     dritten
    2.17
    其妙
    2.14
    Activations Density 0.005%

    No Known Activations