© 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. 34-GEMMASCOPE-RES-131K
    4. 33411
    Prev
    Next
    INDEX
    Explanations

    Man Is

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

    The neuron strongly activates on the successive capitalized words of a news‐style headline, effectively detecting title-case headline tokens.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     my
    -0.98
     flickr
    -0.93
     decreases
    -0.92
     our
    -0.91
     both
    -0.90
     daarna
    -0.89
     more
    -0.89
     Wartawan
    -0.88
     історі
    -0.87
     універ
    -0.87
    POSITIVE LOGITS
     netizens
    0.99
     flexibel
    0.96
     табли
    0.95
     سینم
    0.95
     afla
    0.95
     photos
    0.94
    ጳ
    0.93
    unjukkan
    0.91
     momente
    0.91
    ône
    0.91
    Activations Density 0.011%

    No Known Activations