© 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. 10-GEMMASCOPE-RES-131K
    7. 43367
    Prev
    Next
    INDEX
    Explanations

    seem to

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

    The neuron fires on hedging constructions, specifically occurrences of “seem” or “appear” followed by “to” (e.g. “seem to be,” “appears to have,” “doesn’t appear to allow,” etc.).

    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
     when
    -1.87
     cuando
    -1.65
     should
    -1.45
     These
    -1.43
     until
    -1.41
     могут
    -1.39
     ketika
    -1.38
     Fazit
    -1.36
    즉
    -1.31
     Until
    -1.29
    POSITIVE LOGITS
     hochwertige
    1.41
     entgeg
    1.40
    recently
    1.40
     breite
    1.35
     decorados
    1.32
     cambiando
    1.32
     marcadas
    1.31
     dispuestos
    1.30
     vieilles
    1.30
     ответствен
    1.30
    Activations Density 0.058%

    No Known Activations