© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 72021
    Prev
    Next
    INDEX
    Explanations

    joke endings

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

    The neuron is keyed to the ends of sentences—especially closing punctuation (periods or question marks) and the final word/token just before a sentence boundary.

    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
     Downvote
    -0.99
     Donde
    -0.89
     Idk
    -0.88
     Described
    -0.87
     Ejemplo
    -0.86
     ESPEC
    -0.85
    jelasan
    -0.85
    Especially
    -0.84
    ujące
    -0.83
     Examples
    -0.83
    POSITIVE LOGITS
     joke
    1.65
     jokes
    1.60
     humor
    1.33
     joked
    1.29
     comedy
    1.20
     humour
    1.15
     chistes
    1.13
     satire
    1.05
     joking
    1.05
    wit
    1.05
    Activations Density 0.026%

    No Known Activations