© 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. 71920
    Prev
    Next
    INDEX
    Explanations

    relative pronoun followed by verb

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

    The neuron fires strongly on subordinating or relative-clause markers—words like “who,” “which,” and “that” that introduce clauses.

    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
     rör
    -1.20
    ต่างๆ
    -1.02
    其中
    -1.01
    </h2>
    -1.00
    が
    -1.00
     dieselben
    -0.98
     उस
    -0.97
     diversas
    -0.97
     considerably
    -0.97
    </h4>
    -0.97
    POSITIVE LOGITS
    ようになりました
    1.39
    alamu
    1.30
    chemia
    1.28
    了你
    1.28
    corrhi
    1.22
    kyou
    1.18
     verko
    1.17
    curity
    1.12
    
    1.10
    為に
    1.09
    Activations Density 0.062%

    No Known Activations