© 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. 22-GEMMASCOPE-RES-131K
    7. 66050
    Prev
    Next
    INDEX
    Explanations

    eigenvalues and eigenvectors

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

    The neuron selectively activates on LaTeX‐style math tokens—especially names of matrices and linear‐algebra objects (e.g. T, P, v₁, eigenvectors) and connectors like “and” within mathematical expressions.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     béisbol
    -1.34
     diameter
    -1.15
     boxeo
    -1.08
    การณ์
    -1.05
    Plot
    -1.04
    plot
    -1.03
     "
    -1.01
     sicuro
    -1.00
     deportivas
    -0.97
    konto
    -0.96
    POSITIVE LOGITS
    Camilla
    1.10
    接下来的
    1.08
    Quir
    1.08
    Monarch
    1.07
    Hector
    1.02
    Gente
    1.02
    Jem
    1.01
     Selle
    0.99
    中间
    0.98
     impo
    0.98
    Activations Density 0.009%

    No Known Activations