© 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. 10-GEMMASCOPE-RES-131K
    7. 101603
    Prev
    Next
    INDEX
    Explanations

    entities relative to observer

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

    The neuron detects spatial‐relationship terms that locate people or things relative to a reference (e.g. “behind,” “beside,” “around,” “in front of”).

    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
     is
    -2.05
     according
    -1.62
    1
    -1.57
    了两
    -1.55
    イイ
    -1.46
     हैं
    -1.45
    2
    -1.40
    让她
    -1.38
    がよく
    -1.38
    をした
    -1.38
    POSITIVE LOGITS
    夘
    1.82
    "
    1.68
     leal
    1.61
    n
    1.56
    ｢
    1.55
    。
    1.55
    嗵
    1.53
    HASAN
    1.53
    سایت
    1.51
    𝗪
    1.49
    Activations Density 0.017%

    No Known Activations