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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 40-GEMMASCOPE-2-RES-262K
    7. 4263
    Prev
    Next
    INDEX
    Explanations

    only respond in JSON

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

    sentences or tokens that specify high-level instructions, output format requirements, and schema/JSON constraints (system-level formatting and extraction directives).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    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
    ちょっと
    0.42
     \,
    0.39
    やっぱり
    0.37
     সত্যিই
    0.36
    pretty
    0.34
     সম্ভবত
    0.34
     trochę
    0.34
     やっぱり
    0.33
    clearRect
    0.33
    
    0.33
    POSITIVE LOGITS
     antwort
    0.47
     paraphr
    0.46
     henceforth
    0.45
     paraphrase
    0.45
    のみ
    0.44
     ONLY
    0.44
     वाक्य
    0.43
     sentences
    0.43
     மட்டுமே
    0.43
    不超过
    0.42
    Activations Density 0.525%

    No Known Activations