© 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. 93028
    Prev
    Next
    INDEX
    Explanations

    not happy

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

    The neuron is keyed to negation cues—especially contracted negation endings (like “n’t” or “ain’t”) and explicit negative words (“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
    uró
    -0.94
    ImageContext
    -0.90
    ggiamento
    -0.87
    WHM
    -0.84
     Wählen
    -0.84
    encrypt
    -0.82
     Hvis
    -0.81
    Iné
    -0.80
    itados
    -0.79
    Mime
    -0.78
    POSITIVE LOGITS
     thrilled
    1.44
     liked
    1.28
     impressed
    1.19
     appreciate
    1.14
     pleased
    1.13
     liking
    1.12
     happy
    1.10
     too
    1.10
     appreciated
    1.05
    ชอบ
    1.05
    Activations Density 0.029%

    No Known Activations