© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 335
    Prev
    Next
    INDEX
    Explanations

    nucleic acids

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

    concepts related to absence or lack of something, such as "without money", "without installation", or "no electricity".

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron fires on concrete noun tokens—especially names of “things” or resources (e.g. coding, installation, coursework, word, food)—typically in constructions about having, not having, adding, or removing them.

    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
    orov
    -1.00
    delsen
    -0.98
     verkligen
    -0.98
     שלא
    -0.95
     BaseActivity
    -0.91
     хорошего
    -0.91
     without
    -0.89
     from
    -0.88
    ímica
    -0.87
    volvement
    -0.87
    POSITIVE LOGITS
     nor
    1.01
    任何
    1.00
     stolz
    0.94
     any
    0.94
    ELESS
    0.93
    Therefore
    0.92
     कोई
    0.91
     Instead
    0.91
     כלל
    0.90
    而是
    0.89
    Activations Density 0.078%

    No Known Activations