© 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-3-12B-IT
    3. 12-GEMMASCOPE-2-RES-262K
    4. 3074
    Prev
    Next
    INDEX
    Explanations

    blog **Series** & Video **Series**..."- "...our **series** on..."- "...YouTube video **series**"- "...this **series** has a dedicated section..."- "...workshop **series** to introduce..."- "...(Rick Riordan):** (**Series** Starter)..."- "...podcast **series** about the history..."- "...current **series** where the new **series** began..."The word "series" is consistently present in the activating texts and is the dominant token in `MAX_ACTIVATING_TOKENS`.The explanation should be concise and focus on what the neuron detects.The word "series" is the most obvious pattern.The suggested explanation is "series".It fits the word count and the rules.series

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-it/resid_post/layer_12_width_262k_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
     sedang
    1.78
     pernah
    1.72
    stå
    1.70
     sementara
    1.69
    та
    1.63
     trabalh
    1.63
    म
    1.63
     devolver
    1.61
     chegar
    1.59
    く
    1.54
    POSITIVE LOGITS
    OT
    1.57
    ін
    1.55
    OR
    1.55
    garden
    1.53
    গুপ্ত
    1.52
    ে
    1.48
    ग्राम
    1.48
    }/>
    1.47
    })
    1.44
    cribe
    1.44
    Activations Density 0.017%

    No Known Activations