© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 38317
    Prev
    Next
    INDEX
    Explanations

    and/or followed by adjectives

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

    The neuron activates on words that label specific subtypes or classes (e.g. “public,” “adaptive,” “ordinary,” “secondary,” “chronic,” “inflammatory,” 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_34/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
    を
    -1.31
     نمایید
    -1.21
    が
    -1.20
    など
    -1.18
    밌
    -1.11
    除此之外
    -1.09
    で
    -1.08
     :
    -1.08
    かというと
    -1.07
    があった
    -1.05
    POSITIVE LOGITS
    為に
    1.21
    に入れる
    1.13
    ところに
    1.13
    たびに
    1.04
    salut
    1.01
    umplimiento
    0.99
    hecimento
    0.98
    必要がある
    0.96
    様な
    0.95
    ほうがいい
    0.95
    Activations Density 0.192%

    No Known Activations