© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 2300
    Prev
    Next
    INDEX
    Explanations

    phrases and headings

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

    The neuron strongly activates on evaluative, praise-style adjective and adverb phrases (e.g. “Has very good …,” “more finish,” “great importance”)—i.e. quality-judging language in reviews.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     поскольку
    -1.16
     insofar
    -1.11
    Basic
    -1.10
     sensed
    -1.07
     BASIC
    -1.01
     เน
    -1.00
    basic
    -0.99
     basic
    -0.97
     viewpoint
    -0.97
     envision
    -0.96
    POSITIVE LOGITS
     appliances
    0.99
     anxiously
    0.98
     advertisements
    0.98
     anxiety
    0.96
     females
    0.96
     metaphorical
    0.96
     alterations
    0.90
     equilibrio
    0.89
     vouchers
    0.87
     capabilities
    0.86
    Activations Density 0.023%

    No Known Activations