© 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. 89951
    Prev
    Next
    INDEX
    Explanations

    up and down movement

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

    The neuron activates on the directional adverbs “up” and “down,” i.e. tokens indicating movement or orientation in the vertical direction.

    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
     down
    -1.73
     up
    -1.38
    Down
    -1.29
    DOWN
    -1.27
    down
    -1.20
    ลง
    -1.16
    回去
    -1.02
     Down
    -0.98
    Up
    -0.96
     DOWN
    -0.92
    POSITIVE LOGITS
     to
    2.09
     into
    1.89
     towards
    1.46
     toward
    1.46
     Toward
    1.32
     past
    1.22
    towards
    1.21
     Towards
    1.16
     through
    1.13
    Toward
    1.10
    Activations Density 0.035%

    No Known Activations