© 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. 118203
    Prev
    Next
    INDEX
    Explanations

    competition prize availability

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

    The neuron activates on legal‐style conditional clauses (e.g. “in the event of … being …” or “if a copy of … was not …”).

    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
     lehet
    -1.14
     and
    -0.91
     प्रत्येक
    -0.88
     suelen
    -0.86
     Bearbeit
    -0.85
     mögliche
    -0.85
    遇到的
    -0.84
     недавно
    -0.84
     recently
    -0.81
     môžu
    -0.81
    POSITIVE LOGITS
     becomes
    1.20
     после
    1.10
     any
    1.08
    eafter
    1.05
     herhangi
    1.04
     discover
    1.02
    chafter
    1.00
     discovers
    0.98
     worin
    0.95
     find
    0.95
    Activations Density 0.019%

    No Known Activations