© 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-TRANSCODER-262K
    7. 220299
    Prev
    Next
    INDEX
    Explanations

    circumventing safety protocols

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_31_width_262k_l0_small_affine
    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
     Bes
    0.38
     conformal
    0.37
     Susp
    0.36
     Sol
    0.36
    <!---
    0.35
    paredStatement
    0.35
     Sus
    0.35
    痙
    0.35
     Wetland
    0.34
    ্ভ
    0.34
    POSITIVE LOGITS
     backdoor
    0.68
     sneak
    0.65
     sneaking
    0.63
    Indirect
    0.56
    via
    0.55
     via
    0.54
     indirect
    0.53
     indire
    0.52
     indirectly
    0.51
     Indirect
    0.49
    Activations Density 0.022%

    No Known Activations