© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 40-GEMMASCOPE-2-RES-262K
    7. 8017
    Prev
    Next
    INDEX
    Explanations

    careful consideration

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/resid_post/layer_40_width_262k_l0_medium
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    就可以
    0.37
    ريب
    0.36
    代わりに
    0.36
    Mâc
    0.36
     flound
    0.34
     skyscrapers
    0.33
     podemos
    0.33
    𒊏
    0.33
     შეუ
    0.33
    ప్త
    0.32
    POSITIVE LOGITS
     careful
    1.15
     carefully
    1.10
     beware
    1.10
     Careful
    1.03
    Beware
    1.00
     Carefully
    0.98
     सावधानी
    0.98
     Beware
    0.95
    慎
    0.94
     тщательно
    0.93
    Activations Density 0.104%

    No Known Activations