© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 53951
    Prev
    Next
    INDEX
    Explanations

    said followed by a name

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

    The neuron fires strongly on quoted‐speech attributions—i.e. where someone “said,” “says,” etc., introducing a speaker’s remark.

    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
    遭
    -0.90
     every
    -0.89
     it
    -0.88
     just
    -0.88
    不好
    -0.85
     Just
    -0.84
     all
    -0.84
     first
    -0.83
     impress
    -0.82
     necessità
    -0.82
    POSITIVE LOGITS
     which
    1.49
     ซึ่ง
    1.20
    ซึ่ง
    1.13
    ributes
    1.09
    whose
    1.09
    
    1.09
     cujo
    1.03
     cuya
    1.03
     POSSIBLE
    1.02
     cuyo
    1.00
    Activations Density 0.021%

    No Known Activations