© 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. 59942
    Prev
    Next
    INDEX
    Explanations

    public speeches and events

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

    The neuron is primarily detecting verbs that describe events or actions (especially past‐tense and gerund forms) in the text.

    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
     ilustrasi
    -0.96
    辋
    -0.89
     délai
    -0.89
     ujar
    -0.85
     подска
    -0.84
    𝐈
    -0.84
    Wednesday
    -0.79
    ğini
    -0.79
     Agosto
    -0.77
     séjour
    -0.76
    POSITIVE LOGITS
     speeches
    2.69
     speech
    2.19
     presentation
    1.64
     presentations
    1.63
     brief
    1.63
     Speeches
    1.55
    speech
    1.49
     musical
    1.48
     performances
    1.43
     awards
    1.41
    Activations Density 0.015%

    No Known Activations