© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 65352
    Prev
    Next
    INDEX
    Explanations

    with followed by entity/descriptor

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

    The neuron strongly activates on the preposition “with,” especially when it introduces partners or parties in legal/contractual contexts.

    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
     apropri
    -1.36
     (_.
    -1.30
    も含
    -1.29
     reden
    -1.27
    を発見
    -1.25
    を含
    -1.21
    GRUPO
    -1.21
     enfant
    -1.20
     utilis
    -1.20
     svou
    -1.18
    POSITIVE LOGITS
     for
    1.70
    koľ
    1.48
    对于
    1.47
     When
    1.45
     by
    1.37
     Unfortunately
    1.37
     Generally
    1.36
     jeweiligen
    1.34
     Also
    1.34
     Regarding
    1.34
    Activations Density 0.054%

    No Known Activations