© 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-3-27B-IT
    5. 31-GEMMASCOPE-2-RES-262K
    7. 12145
    Prev
    Next
    INDEX
    Explanations

    model responses

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

    the beginning of the assistant’s reply in a dialogue (the assistant turn marker or first token of the model’s message).

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/resid_post/layer_31_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
     regressions
    0.47
     opérations
    0.46
     വിൽപ്പന
    0.45
     variété
    0.45
     auteurs
    0.45
     radiographs
    0.45
    ÜR
    0.44
    鸮
    0.44
     ফ্যাস
    0.43
     collaborateurs
    0.42
    POSITIVE LOGITS
     sorry
    0.69
    sorry
    0.69
     plz
    0.54
    Sorry
    0.49
     Sorry
    0.49
    hehe
    0.49
     pls
    0.47
    they
    0.47
    answer
    0.46
    you
    0.46
    Activations Density 0.007%

    No Known Activations