© 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. 22-GEMMASCOPE-RES-131K
    4. 39305
    Prev
    Next
    INDEX
    Explanations

    locations, displacement, relative

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

    The neuron is primarily triggered by technical/scientific tokens—especially numbers and domain‐specific terminology—typical of research paper text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     progressBar
    -1.01
    VDC
    -0.97
    óny
    -0.96
     membuk
    -0.94
     ricordare
    -0.94
     partecipare
    -0.93
     attivo
    -0.93
     condividere
    -0.92
     овощи
    -0.92
    хие
    -0.92
    POSITIVE LOGITS
     chegada
    1.13
    都需要
    1.11
     châ
    1.09
    ]})
    1.05
    妪
    1.03
    考验
    1.02
    örgy
    1.02
    殃
    1.02
     sabes
    1.01
     tierras
    1.00
    Activations Density 0.012%

    No Known Activations