© 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-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 73986
    Prev
    Next
    INDEX
    Explanations

    software setup

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

    The neuron fires on standout header or title tokens—strong activations on single, often capitalized or punctuated words (and numbers) you’d see in product, download, or ad headlines.

    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
    Filmografie
    -1.10
    estinations
    -1.05
    ENODEV
    -1.04
    coronavirus
    -1.00
     想
    -1.00
    abler
    -1.00
    nehm
    -0.98
    האם
    -0.97
    Več
    -0.97
     δεν
    -0.96
    POSITIVE LOGITS
    あー
    1.16
    やっぱ
    1.13
     människor
    0.98
     that
    0.97
     pesso
    0.96
    の詳細
    0.94
     podría
    0.94
     социальных
    0.93
    うーん
    0.91
    来週
    0.91
    Activations Density 0.010%

    No Known Activations