© 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. 113319
    Prev
    Next
    INDEX
    Explanations

    finding solutions

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

    The neuron flags words that describe achieving, solving, or accomplishing something (e.g. “accomplish,” “solved,” “major,” “answer,” “realize,” “solution”).

    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
    sorted
    -1.17
    such
    -1.14
    pides
    -1.07
    scanner
    -1.03
     그리고
    -1.02
     \%$,
    -1.02
    Sí
    -1.00
     such
    -0.98
     thúc
    -0.97
    琥
    -0.97
    POSITIVE LOGITS
    obacz
    1.23
    decisions
    1.13
     hotelu
    1.12
    𓂀
    1.11
     تريد
    1.09
    ཿ
    1.09
    ༒
    1.09
    problemas
    1.08
    Molto
    1.08
    Tamaño
    1.07
    Activations Density 0.362%

    No Known Activations