© 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. 22-GEMMASCOPE-RES-131K
    4. 9319
    Prev
    Next
    INDEX
    Explanations

    engine trouble and duration

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

    This neuron activates on sentences that describe technical troubleshooting of a device or vehicle—i.e. detailed problem‐statement narratives (e.g., “I will start it and the engine would turn over but when I let go of the key the car would die”).

    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
     your
    -1.55
     seçenek
    -1.34
     rappresenta
    -1.30
     you
    -1.30
     çünkü
    -1.30
     todėl
    -1.26
    歟
    -1.26
     fırs
    -1.26
    合作
    -1.25
    べし
    -1.23
    POSITIVE LOGITS
     but
    1.69
     seemed
    1.38
    ++);
    1.31
     بازیگر
    1.27
    capulco
    1.24
     سریال
    1.23
     १
    1.23
     音乐
    1.23
    ༼
    1.21
    seem
    1.17
    Activations Density 0.085%

    No Known Activations