© 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. 34-GEMMASCOPE-RES-131K
    4. 45351
    Prev
    Next
    INDEX
    Explanations

    numbers and probabilities

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

    The neuron is a “long‐token” detector—it lights up on unusually long words or identifiers (e.g. full technical terms or code names) rather than on short/common subwords.

    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
     nevertheless
    -1.01
     however
    -0.90
     nonetheless
    -0.80
    便利な
    -0.77
    ㈱
    -0.74
     ہم
    -0.73
     Uttarakhand
    -0.71
     χρό
    -0.71
    甚至是
    -0.71
     notwithstanding
    -0.71
    POSITIVE LOGITS
    Chile
    1.31
     Chilean
    1.24
    Yam
    1.07
     Yam
    1.06
    Portland
    1.05
    chile
    0.97
     Chile
    0.94
     Portland
    0.92
    JVM
    0.88
    Oregon
    0.85
    Activations Density 0.000%

    No Known Activations