© 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
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 93509
    Prev
    Next
    INDEX
    Explanations

    how something occurs

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

    The neuron is specifically detecting the adverb “somehow.”

    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
     sarebbe
    -1.01
     frankly
    -1.01
    مقدمه
    -1.00
     tikra
    -0.99
    λίου
    -0.98
     solcher
    -0.97
     infância
    -0.97
    pompa
    -0.96
     qualsiasi
    -0.95
     constater
    -0.95
    POSITIVE LOGITS
     or
    1.78
     managed
    1.45
     manage
    1.45
     manages
    1.43
     still
    1.23
     some
    1.20
     related
    1.18
     managing
    1.17
     Manage
    1.15
     magically
    1.14
    Activations Density 0.016%

    No Known Activations