© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 26682
    Prev
    Next
    INDEX
    Explanations

    when is

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

    The neuron flags mentions of the narrative’s agent or subject—proper‐name characters or first‐person pronouns (I/we) that indicate who is doing the action.

    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
     Implant
    -1.07
     手绘
    -1.05
    Введение
    -1.04
    雖然
    -1.04
     火鍋
    -1.03
     排骨
    -1.02
     homoto
    -1.02
    Хотя
    -1.00
     historial
    -0.99
     内衣
    -0.98
    POSITIVE LOGITS
     asked
    1.26
     faced
    1.21
     pierwszy
    1.18
     chegar
    1.10
     finally
    1.09
     I
    1.09
     viewed
    1.09
    わけではない
    1.09
     chegou
    1.09
     arrived
    1.07
    Activations Density 0.101%

    No Known Activations