© 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. 117719
    Prev
    Next
    INDEX
    Explanations

    network addresses and headers

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

    This neuron fires on network‐address–related tokens (e.g. “inet,” “in,” “addr,” “ip”) in code.

    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
    allenges
    -0.75
    machinery
    -0.74
    nium
    -0.73
    
    -0.73
     прогно
    -0.73
     ایک
    -0.72
     due
    -0.71
    陷阱
    -0.71
    Pile
    -0.70
    Work
    -0.69
    POSITIVE LOGITS
    зер
    0.90
     SIN
    0.86
     Ina
    0.78
     ~$
    0.77
     saith
    0.77
     Dars
    0.77
     babel
    0.76
     Interceptor
    0.76
     Quarantine
    0.75
    sin
    0.75
    Activations Density 0.006%

    No Known Activations