© 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. 44438
    Prev
    Next
    INDEX
    Explanations

    gaining understanding and making realizations

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

    The neuron flags expressions of personal learning or realization—especially verbs like “learned,” “know,” “discovered,” or “come to (believe/know)” that introduce what the speaker has come to understand.

    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
     its
    -1.59
     his
    -1.42
    ↵↵
    -1.41
     at
    -1.23
     her
    -1.23
     La
    -1.19
     indoctr
    -1.12
    菓
    -1.10
     The
    -1.09
    密封
    -1.09
    POSITIVE LOGITS
     pegatina
    1.59
     macetas
    1.56
     also
    1.56
     tunik
    1.52
     sudadera
    1.50
     rằng
    1.50
     récomp
    1.48
     phénom
    1.45
    蒌
    1.42
     llavero
    1.41
    Activations Density 0.039%

    No Known Activations