Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Gemma-3-27B-IT
    3. 31-GEMMASCOPE-2-RES-16K
    4. 12065
    Prev
    Next
    INDEX
    Explanations

    model response start

    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_31_width_16k_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
    훙
    0.24
     イタリア
    0.24
    ックレス
    0.24
     exudes
    0.23
     señor
    0.22
     यांच्या
    0.22
     nouvel
    0.21
    ρώ
    0.21
    వారి
    0.21
    賀
    0.21
    POSITIVE LOGITS
    因为
    0.33
     చాలా
    0.30
     కానీ
    0.30
     Ideally
    0.30
    将
    0.29
     Think
    0.29
    अधिकांश
    0.29
     Якщо
    0.28
     বেশিরভাগ
    0.27
    এখানে
    0.27
    Activations Density 0.031%

    No Known Activations