© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 38684
    Prev
    Next
    INDEX
    Explanations

    pronouns and general statements

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

    The neuron responds strongly to common high-frequency function words (especially pronouns and auxiliary verbs such as “it,” “is,” “they,” “have,” “been,” 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_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
    bols
    -1.18
     dégustation
    -1.18
    对此
    -1.07
    SNMP
    -1.05
    рово
    -1.05
    °°
    -1.05
     Parece
    -1.02
    がございます
    -1.01
    的那种
    -1.01
     Gera
    -1.00
    POSITIVE LOGITS
     any
    1.45
     anything
    1.40
    ↵↵↵
    1.28
     ourselves
    1.24
     myself
    1.23
    ↵↵↵↵↵↵↵
    1.19
     actually
    1.13
    ÁB
    1.11
     hanno
    1.09
    ↵↵↵↵↵
    1.09
    Activations Density 0.233%

    No Known Activations