© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 72278
    Prev
    Next
    INDEX
    Explanations

    compound nouns after a word

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

    This neuron responds to isolated technical identifiers and named entities—e.g. code symbols, acronyms, chemical names, class or variable names, and other proper‐noun–style tokens.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     didnt
    -1.06
     without
    -0.98
     before
    -0.96
     नहीं
    -0.94
    entukan
    -0.94
     because
    -0.92
     only
    -0.92
     first
    -0.92
     initially
    -0.88
     نہیں
    -0.88
    POSITIVE LOGITS
     acrobat
    0.97
    ′
    0.94
    },
    0.88
     ônibus
    0.87
     zufolge
    0.87
    }');
    0.85
     разных
    0.84
    }.
    0.84
    ]}>
    0.84
    ″
    0.82
    Activations Density 0.117%

    No Known Activations