© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 109237
    Prev
    Next
    INDEX
    Explanations

    jwt algorithms

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

    The neuron is strongly activated by code‐style tokens—things like acronyms (JWT, ALG, NONE), version numbers, numeric literals, and underscore‐separated identifiers.

    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
     Dimensional
    -0.77
     Bura
    -0.75
    佐
    -0.70
     petri
    -0.69
    леб
    -0.68
     полови
    -0.68
     portail
    -0.68
    crom
    -0.67
    までに
    -0.67
     camo
    -0.67
    POSITIVE LOGITS
     alg
    1.09
    alg
    1.05
     typ
    1.05
     kid
    1.01
    kid
    1.00
    Alg
    0.99
     Kid
    0.95
     HS
    0.94
    typ
    0.93
    claims
    0.89
    Activations Density 0.025%

    No Known Activations