© 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. 34680
    Prev
    Next
    INDEX
    Explanations

    in code or as a preposition

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

    The neuron strongly activates on the token “in” (particularly as the keyword or identifier “in” in code).

    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
     when
    -1.39
    
    -1.30
     or
    -1.26
     if
    -1.15
     only
    -1.05
     during
    -1.02
     kecelakaan
    -0.99
    الى
    -0.98
     ONLY
    -0.98
     this
    -0.98
    POSITIVE LOGITS
     delightfully
    1.05
     inſ
    1.04
     myriad
    1.03
     dobla
    1.02
    endido
    1.01
    nischer
    1.00
     adorned
    0.98
     البرنامج
    0.98
     Fernsehen
    0.96
     ludicrous
    0.95
    Activations Density 0.001%

    No Known Activations