© 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-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 7402
    Prev
    Next
    INDEX
    Explanations

    waste of or abuse of

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron detects terms denoting misuse or wrongdoing (e.g. waste, abuse, breach).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    氿
    -2.09
    鸸
    -2.06
    ఽ
    -2.03
    鸲
    -2.02
    乄
    -1.94
     vinte
    -1.91
    皤
    -1.79
    меди
    -1.79
     simpl
    -1.77
    ⚋
    -1.74
    POSITIVE LOGITS
     and
    2.20
     enormous
    2.14
     perhaps
    2.13
     or
    2.06
     has
    1.99
    Before
    1.94
     also
    1.92
     enormously
    1.87
     sometimes
    1.86
     considerable
    1.84
    Activations Density 0.045%

    No Known Activations