© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 99226
    Prev
    Next
    INDEX
    Explanations

    Q: or Dyspnea

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

    Japanese language text and technical terms.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron responds to subordinating or conditional/temporal connectives (e.g. Japanese ときに, ~があると, ~ても, ところ, ~だと, ~か and English “implies,” etc.) that introduce clauses.

    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
     also
    -2.28
     as
    -2.27
     with
    -1.63
     for
    -1.60
     one
    -1.51
     actually
    -1.48
     just
    -1.48
     a
    -1.45
    шт
    -1.43
     so
    -1.42
    POSITIVE LOGITS
     diriger
    1.43
    This
    1.41
     Permite
    1.36
     nozze
    1.35
    manera
    1.33
    </h2>
    1.32
     canale
    1.30
     Utiliza
    1.30
     scènes
    1.29
     Änder
    1.28
    Activations Density 0.002%

    No Known Activations