© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 31-GEMMASCOPE-2-RES-262K
    7. 90408
    Prev
    Next
    INDEX
    Explanations

    conflictingcontrasting"on one hand"

    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
     redund
    0.44
     ANY
    0.42
    人も
    0.41
     basics
    0.41
     次
    0.40
     YOUR
    0.40
     আরও
    0.40
     any
    0.39
     మరింత
    0.38
     FROM
    0.38
    POSITIVE LOGITS
    虽然
    0.65
    雖然
    0.63
     zwar
    0.59
    雖
    0.58
     conflicting
    0.56
    while
    0.55
    虽
    0.54
     paradoxical
    0.53
    一方面
    0.52
     contradictory
    0.52
    Activations Density 0.220%

    No Known Activations