© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 6685
    Prev
    Next
    INDEX
    Explanations

    pleasant experiences and conditions

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

    The neuron fires strongly on the word “pleasant” (and very weakly on similar positive descriptors), i.e. it detects that specific positive adjective.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     magnificent
    -2.28
    肪
    -2.05
     marvelous
    -1.97
    紝
    -1.93
    迓
    -1.92
     splendid
    -1.89
     envía
    -1.85
     fabulous
    -1.82
     aperto
    -1.80
     genesis
    -1.78
    POSITIVE LOGITS
    }
    2.02
     एवं
    1.80
    ского
    1.77
     Using
    1.77
     Also
    1.77
     que
    1.73
    Even
    1.70
     Even
    1.70
     That
    1.70
     Taking
    1.66
    Activations Density 0.007%

    No Known Activations