© 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. 34-GEMMASCOPE-RES-131K
    7. 63735
    Prev
    Next
    INDEX
    Explanations

    found harmed or dead

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

    The neuron flags news‐style “found” statements describing victims or bodies (e.g. found, crumpled, dead, stabbed, unresponsive).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     khiến
    -0.91
     hueso
    -0.89
    NSMutableArray
    -0.86
     proporcion
    -0.85
     incó
    -0.84
    也就
    -0.84
     sturdy
    -0.82
     year
    -0.81
    怖い
    -0.81
     poursuit
    -0.81
    POSITIVE LOGITS
     lying
    2.30
    Lying
    1.69
     bleeding
    1.53
     Lying
    1.53
     lies
    1.52
     unconscious
    1.48
     laying
    1.45
     motionless
    1.41
     лежит
    1.38
     lifeless
    1.35
    Activations Density 0.028%

    No Known Activations