© 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. 38850
    Prev
    Next
    INDEX
    Explanations

    "there has to be"

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

    This neuron detects content words inside quoted internal thoughts (i.e. the words in the “I thought, ‘…’” passages).

    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
     your
    -1.81
     our
    -1.38
     on
    -1.38
     répondu
    -1.27
     YOUR
    -1.24
     toreb
    -1.22
    these
    -1.22
     you
    -1.21
    ↵
    -1.20
     Unterstüt
    -1.14
    POSITIVE LOGITS
    ]:
    1.47
    AdvEx
    1.41
     magnitud
    1.38
    othermic
    1.22
    ‌پدی
    1.20
    ".
    1.20
     /\.(
    1.20
    cended
    1.19
    </h1>
    1.18
    haviours
    1.17
    Activations Density 0.029%

    No Known Activations