© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 27069
    Prev
    Next
    INDEX
    Explanations

    named in honor of

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

    This neuron fires on geographic place-type words and naming contexts (e.g. “town,” “city,” “named,” “Peak,” “Bay,” “community,” 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_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
    })}$
    -0.96
    indicated
    -0.93
     nonatomic
    -0.89
     хуже
    -0.89
    ugna
    -0.88
     CUTE
    -0.88
    とにかく
    -0.87
    ที่จะ
    -0.85
    قديم
    -0.84
     asisti
    -0.84
    POSITIVE LOGITS
     his
    1.47
     namesake
    1.27
     dedicated
    1.21
     named
    1.21
     name
    1.13
     honour
    1.09
    寤
    1.03
     그의
    1.02
    近くに
    0.98
     commemorate
    0.97
    Activations Density 0.040%

    No Known Activations