© 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. 20612
    Prev
    Next
    INDEX
    Explanations

    specialization creates blind spots

    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
     各
    0.52
     Bull
    0.46
     vanwege
    0.44
     nakk
    0.44
     School
    0.43
     将
    0.43
    wissenschaft
    0.43
     Governing
    0.43
     Framework
    0.43
     نوٹ
    0.42
    POSITIVE LOGITS
    ştir
    0.47
    nnnn
    0.45
    也許
    0.44
    ১
    0.41
    ogy
    0.41
    gore
    0.41
    ă
    0.41
    тель
    0.40
    слен
    0.40
    㥽
    0.40
    Activations Density 0.000%

    No Known Activations