© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 66877
    Prev
    Next
    INDEX
    Explanations

    product qualities and features

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

    The neuron detects product feature and benefit terms in marketing or technical descriptions (e.g. comfort, durability, protection, breathability, fit, etc.).

    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
     their
    -1.43
     then
    -1.34
     it
    -1.20
     how
    -1.16
    に行ってきました
    -1.16
     If
    -1.16
    の形
    -1.10
     daarvoor
    -1.08
     Obtén
    -1.05
     Entonces
    -1.05
    POSITIVE LOGITS
     your
    1.34
     and
    1.26
     هنگام
    1.20
     also
    1.19
     you
    1.15
     podczas
    1.13
     WORKSHOP
    1.09
    ьез
    1.08
     durante
    1.06
    ussch
    1.05
    Activations Density 0.036%

    No Known Activations