© 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. 24-GEMMASCOPE-2-RES-16K
    7. 12088
    Prev
    Next
    INDEX
    Explanations

    The neuron specializes in spotting technical method‐step language—particularly claim-style action verbs and “step of” phrasing that enumerate or describe procedural steps.

    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_24_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
    backed
    0.75
     versions
    0.74
    m
    0.71
    lardan
    0.67
     ürün
    0.66
     цін
    0.65
    d
    0.64
     považ
    0.64
     Faber
    0.64
    v
    0.63
    POSITIVE LOGITS
    初始化
    1.16
     실행
    1.15
     初始化
    1.10
    initialize
    1.09
     initialize
    1.03
    调用
    1.01
     Initialize
    0.99
     componentDid
    0.99
     ngOnInit
    0.96
    execute
    0.96
    Activations Density 1.424%

    No Known Activations