© 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. 10-GEMMASCOPE-RES-131K
    4. 57427
    Prev
    Next
    INDEX
    Explanations

    when I

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

    The neuron specifically fires on the subordinating conjunction “When,” marking the start of temporal or conditional 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_10/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
    in
    -2.53
    is
    -2.53
    {
    -2.30
    de
    -2.19
    ally
    -2.14
    』、
    -2.09
    </h3>
    -2.06
    m
    -2.05
     has
    -2.03
    er
    -2.00
    POSITIVE LOGITS
    नलोड
    2.47
    providedIn
    2.19
     }));
    2.17
    </b>
    2.16
    大多
    2.13
     They
    2.11
    EDEFAULT
    2.08
    ࿚
    2.05
     If
    2.03
     Therefore
    2.00
    Activations Density 0.031%

    No Known Activations