© 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. 46427
    Prev
    Next
    INDEX
    Explanations

    breakdown, analysis, explanation

    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_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
     R
    0.51
    7
    0.47
     O
    0.47
     E
    0.45
    6
    0.45
    0
    0.44
     very
    0.42
    2
    0.42
     highly
    0.41
     N
    0.41
    POSITIVE LOGITS
    และการ
    0.50
    owaniem
    0.46
     иных
    0.43
    semblance
    0.42
    侵害
    0.41
     рэгістрацыі
    0.40
    oration
    0.40
    이를
    0.39
    溲
    0.39
    Miranda
    0.39
    Activations Density 0.053%

    No Known Activations