© 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. 22-GEMMASCOPE-RES-131K
    7. 73342
    Prev
    Next
    INDEX
    Explanations

    object state changes

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

    The neuron strongly activates on technical terms related to data‐persistence and object/context management (e.g. “managed,” “persistent,” “context,” “objects,” “merge,” “unload,” 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_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
     poë
    -1.16
     INSTALLATION
    -1.10
     sains
    -1.09
     tfsi
    -1.09
     stockage
    -1.09
     interieur
    -1.09
     parfums
    -1.08
     tabac
    -1.07
     nabi
    -1.05
     salade
    -1.05
    POSITIVE LOGITS
     necessary
    1.15
     what
    1.09
    Several
    1.09
     several
    1.07
     changed
    1.05
    Many
    1.04
    Some
    1.03
     before
    1.02
     then
    1.02
     buvo
    0.99
    Activations Density 0.004%

    No Known Activations