© 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. 40-GEMMASCOPE-2-RES-262K
    7. 137089
    Prev
    Next
    INDEX
    Explanations

    attention, control characters, Minister

    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
    Das
    0.39
     asing
    0.37
    Established
    0.37
     এতদিন
    0.37
     najbli
    0.36
    Property
    0.36
    ");
    0.36
     lhe
    0.35
     }
    0.35
    oves
    0.35
    POSITIVE LOGITS
    deleteOne
    0.42
    狒
    0.40
    रांची
    0.39
    夭
    0.38
    ahal
    0.38
    不是很
    0.37
     consonants
    0.37
    消失
    0.36
     عليكم
    0.36
    结婚
    0.36
    Activations Density 0.001%

    No Known Activations