© 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-3-12B
    5. 41-GEMMASCOPE-2-RES-65K
    7. 13154
    Prev
    Next
    INDEX
    Explanations

    apologize or sorry

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_41_width_65k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    వంత
    0.55
     Rating
    0.52
     benef
    0.51
     rating
    0.51
     Realistic
    0.51
     CLIN
    0.50
     Fitch
    0.49
     discriminant
    0.48
     Invis
    0.48
     ratings
    0.48
    POSITIVE LOGITS
     apologies
    4.28
     apologize
    4.22
     apology
    4.19
     sorry
    4.16
     Sorry
    4.03
    sorry
    4.00
     apologized
    3.97
     apologise
    3.97
    Sorry
    3.92
     apologizing
    3.89
    Activations Density 0.138%

    No Known Activations