© 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. 14921
    Prev
    Next
    INDEX
    Explanations

    narrow and tight

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

    The neuron detects words describing constriction or limited width—i.e. terms that convey something being narrow or tight.

    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
    ické
    -1.08
    揶
    -1.06
     pôr
    -1.06
     Ensino
    -1.05
     open
    -1.00
    新潟
    -1.00
     governador
    -0.97
     chefe
    -0.96
     prefeito
    -0.94
     relax
    -0.94
    POSITIVE LOGITS
     tight
    1.91
     tightly
    1.85
    tight
    1.76
     tighter
    1.59
    Tight
    1.55
     closed
    1.42
     Tight
    1.41
     narrow
    1.39
     tightened
    1.35
     tightness
    1.34
    Activations Density 0.110%

    No Known Activations