© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 15873
    Prev
    Next
    INDEX
    Explanations

    memory and remembering

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

    The neuron fires strongly on first-person reflections about recalling or preserving memories—especially around organizing photos, writing captions, and “remembering” past events.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    bucks
    0.51
     Beled
    0.49
     proxy
    0.48
    🕊
    0.48
     ఇం
    0.47
    ниците
    0.47
     Nxc
    0.46
    ruiting
    0.46
    璇
    0.46
    ♘
    0.46
    POSITIVE LOGITS
     memories
    3.91
     기억
    3.69
     memory
    3.55
     воспомина
    3.55
     Memories
    3.45
    記憶
    3.44
     remembered
    3.41
     remembers
    3.38
     remember
    3.34
    记忆
    3.34
    Activations Density 0.515%

    No Known Activations