© 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. 14755
    Prev
    Next
    INDEX
    Explanations

    http, UNITED, COLOR

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

    The neuron strongly activates on opaque technical identifiers—long alphanumeric tokens such as camel-cased event or class names, version strings, hex hashes, docket numbers, 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
     POSTAGE
    -1.19
     Février
    -1.11
    IDOS
    -1.11
    感が
    -1.10
    EdgeInsets
    -1.09
     Tél
    -1.07
    🤟
    -1.06
     dienst
    -1.04
     lavable
    -1.04
    anmar
    -1.03
    POSITIVE LOGITS
    Recept
    1.23
    Zé
    1.21
    pontos
    1.11
    gettyimages
    1.09
    nhung
    1.06
    Passe
    1.05
     everything
    1.05
    motorola
    1.04
    Seni
    1.03
    conjunto
    1.02
    Activations Density 0.001%

    No Known Activations