© 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. 10-GEMMASCOPE-RES-131K
    7. 7867
    Prev
    Next
    INDEX
    Explanations

    problems and difficulties

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

    The neuron activates on occurrences of the word “problem” (and its plural “problems”), marking mentions of difficulties or issues.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     nevera
    -1.67
     entidade
    -1.59
    謖
    -1.58
     Conſ
    -1.56
     découvert
    -1.56
     soyez
    -1.55
    嶌
    -1.53
     sudadera
    -1.51
     choisi
    -1.50
     attendu
    -1.48
    POSITIVE LOGITS
    P
    1.80
     difficult
    1.67
     even
    1.63
     jika
    1.54
     if
    1.52
    C
    1.52
     difficulties
    1.51
    Surprisingly
    1.48
     all
    1.47
     appear
    1.45
    Activations Density 0.030%

    No Known Activations