© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 25479
    Prev
    Next
    INDEX
    Explanations

    time-related words and temporal markers

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

    This neuron strongly activates on the headline or title region of a post (e.g. product‐release or news‐style announcement text).

    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
    😜
    -1.34
     exis
    -1.17
    }
    -1.15
    вало
    -1.11
    恚
    -1.07
     InputMaybe
    -1.07
     .............
    -1.05
    🤯
    -1.05
    dan
    -1.05
    */
    -1.05
    POSITIVE LOGITS
     today
    2.09
     yesterday
    1.84
     tonight
    1.78
     tomorrow
    1.73
     they
    1.63
     now
    1.59
     apparently
    1.40
     они
    1.30
     bugün
    1.22
    Yesterday
    1.20
    Activations Density 0.086%

    No Known Activations