© 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. 10-GEMMASCOPE-RES-131K
    4. 67646
    Prev
    Next
    INDEX
    Explanations

    events and incidents

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

    The neuron strongly activates on tokens that are clearly not ordinary English words—i.e. foreign‐language or non‐English terms.

    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
     sensação
    -2.38
     resolutely
    -2.22
     страны
    -2.19
    顺便
    -2.16
    だけではなく
    -2.14
    だけでなく
    -2.13
     Hinblick
    -2.13
     малень
    -2.13
    今回も
    -2.11
    r
    -2.11
    POSITIVE LOGITS
    )
    2.78
    ")
    2.64
    ﻒ
    2.63
    ]);
    2.58
     and
    2.50
    .
    2.47
     Of
    2.42
    ],
    2.34
    )((
    2.27
     })
    2.23
    Activations Density 0.000%

    No Known Activations