Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Gemma-3-12B-IT
    3. 24-GEMMASCOPE-2-RES-16K
    4. 154
    Prev
    Next
    INDEX
    Explanations

    The neuron selectively activates on long, multi-syllable, domain-specific technical terms and jargon.

    oai_token-act-pair · o4-miniTriggered by @camilab

    It activates on uncommon/rare or domain-specific tokens — long multi-subword pieces like technical terms, proper nouns, or oddly segmented words.

    oai_token-act-pair · gpt-5-miniTriggered by @camilab
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-it/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     impor
    0.21
     docile
    0.21
     aloof
    0.20
     alterna
    0.20
     shadowy
    0.20
    tenir
    0.20
    λύ
    0.20
     emphatic
    0.20
     assent
    0.19
     amiable
    0.19
    POSITIVE LOGITS
    这个
    0.30
    테
    0.25
    А
    0.24
    ка
    0.24
    他的
    0.24
    е
    0.24
    ا
    0.23
    我对
    0.22
    Ко
    0.22
    這個
    0.22
    Activations Density 1.068%

    No Known Activations