© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 34646
    Prev
    Next
    INDEX
    Explanations

    not necessarily or equal

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

    The neuron activates on statements expressing that items (e.g. files) do not follow a particular naming or formatting convention.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     instead
    -1.23
     chứ
    -1.07
     rather
    -1.05
     вместо
    -1.00
     konfigura
    -0.98
     coloque
    -0.98
     zamiast
    -0.98
     ราคา
    -0.98
     riuscito
    -0.96
    がない
    -0.95
    POSITIVE LOGITS
     necessarily
    1.10
     accurate
    1.10
     familiar
    1.09
     same
    1.03
     exact
    1.01
     anymore
    0.99
     cùng
    0.98
     what
    0.97
     Furthermore
    0.96
     Lieder
    0.94
    Activations Density 0.074%

    No Known Activations