© 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. 21791
    Prev
    Next
    INDEX
    Explanations

    on the

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

    The neuron fires on explicit date/time expressions (e.g. calendar dates and timestamps).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    prepositional uses of “on,” typically following a verb or pronoun to introduce a noun phrase indicating position or involvement.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    the preposition "on" when it appears in common phrases or contexts.

    oai_token-act-pair · claude-4-5-sonnetTriggered 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
    0
    -2.20
     اینکه
    -1.77
    廡
    -1.76
     powiedzieć
    -1.75
    to
    -1.72
     zijne
    -1.70
    3
    -1.69
    1
    -1.63
     hunne
    -1.60
    8
    -1.60
    POSITIVE LOGITS
     a
    2.27
    }),
    
    1.76
     out
    1.75
     escena
    1.60
     "'
    1.57
    僕は
    1.56
    исленность
    1.53
    StoryboardSegue
    1.52
     identifies
    1.52
     instructed
    1.52
    Activations Density 0.035%

    No Known Activations