© 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. 7008
    Prev
    Next
    INDEX
    Explanations

    turning left or right

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

    The neuron specializes in spotting navigation‐style directional commands—verbs that tell you to change course (e.g. “turn,” “split off,” “duck into,” 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_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
    HUR
    -0.91
     chiare
    -0.88
     gesellschaft
    -0.87
    стный
    -0.81
     fidé
    -0.80
    Hep
    -0.79
     rembourse
    -0.79
     美術
    -0.79
    σθαι
    -0.78
    ITAN
    -0.78
    POSITIVE LOGITS
     left
    2.38
     right
    1.74
    left
    1.62
     LEFT
    1.37
     Left
    1.29
    LEFT
    1.29
    Left
    1.28
     at
    1.27
     onto
    1.23
     turn
    1.19
    Activations Density 0.008%

    No Known Activations