© 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
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 280
    Prev
    Next
    INDEX
    Explanations

    directional prepositions and adverbs

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

    The neuron detects numeric location or distance expressions (e.g. “further east,” “several more,” “1.2 km,” “miles up the road”).

    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
     verticale
    -0.97
     vertically
    -0.93
    垂直
    -0.90
    across
    -0.90
    焢
    -0.89
     sideways
    -0.87
     horizontally
    -0.85
     iscri
    -0.84
    WEET
    -0.82
     desmotiv
    -0.82
    POSITIVE LOGITS
     past
    1.46
     distal
    1.33
     towards
    1.26
     proximal
    1.20
     upstream
    1.14
     Past
    1.08
     near
    1.02
     downstream
    1.02
     from
    0.98
     at
    0.93
    Activations Density 0.071%

    No Known Activations