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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 40-GEMMASCOPE-2-RES-262K
    7. 17902
    Prev
    Next
    INDEX
    Explanations

    awful, horrid, terrible, dreadful

    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_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
    λικό
    0.42
    rici
    0.42
     seamless
    0.42
    jaer
    0.41
    inkan
    0.39
    擇
    0.38
    aley
    0.38
     substantially
    0.37
    ureux
    0.37
    transpose
    0.36
    POSITIVE LOGITS
     silly
    0.53
     dreadful
    0.50
     awkward
    0.50
     horrid
    0.47
     embarrassing
    0.45
    瘤
    0.45
     shameful
    0.44
    RAS
    0.44
     wretched
    0.44
     затруд
    0.43
    Activations Density 0.001%

    No Known Activations