© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 130527
    Prev
    Next
    INDEX
    Explanations

    The neuron consistently fires on mid- to high-importance content words in sports reporting—terms that point to key events or stakes (chance, opportunity, track, victory, tournament, squad, weeks, soil, etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    future prospects and opportunities

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    sometimes
    -1.25
     parfois
    -1.14
     jossa
    -1.09
    !&
    -1.05
    almost
    -1.03
     souvent
    -1.01
     plupart
    -1.00
     dokonce
    -0.99
     灰色
    -0.99
     faktisk
    -0.99
    POSITIVE LOGITS
     this
    1.43
     in
    1.40
     next
    1.19
     now
    1.14
     tejto
    1.10
     will
    1.09
     if
    1.06
     tohto
    1.02
     from
    1.00
     tomorrow
    0.98
    Activations Density 0.028%

    No Known Activations