© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 30978
    Prev
    Next
    INDEX
    Explanations

    multi-target tracking

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

    The neuron strongly responds to specialized technical or method names—especially capitalized terms—typical of scientific writing.

    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
    isometric
    -0.81
    𝙝
    -0.77
    noli
    -0.73
     Megan
    -0.68
    厳選
    -0.68
    どこに
    -0.67
    搅拌
    -0.66
    装着
    -0.66
     pory
    -0.66
     intervalo
    -0.65
    POSITIVE LOGITS
     tracks
    1.56
    Track
    1.45
     track
    1.44
     Tracks
    1.44
     tracking
    1.43
     tracker
    1.43
     Track
    1.41
    tracks
    1.41
    TRACK
    1.40
     TRACK
    1.38
    Activations Density 0.033%

    No Known Activations