© 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
    2. Gemma-3-27B
    3. 31-GEMMASCOPE-2-RES-262K
    4. 6319
    Prev
    Next
    INDEX
    Explanations

    beyond"}, {"name": "rise", "context": "Fundrise"}, {"name": "nation", "context": "the nation"}, {"name": "area", "context": "washroom space area"}, {"name": "Wind", "context": "Mountbatten-Windsor"}]This neuron seems to be involved in identifying sequences where a certain positive token is followed by another token that completes a phrase or compound word, often indicating scope, relation, or extension. above and beyond fundraiser nation area wind

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

    numbers and punctuation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-pt/resid_post/layer_31_width_262k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     benutzen
    0.27
     haha
    0.27
     Haha
    0.27
     veldig
    0.26
     すご
    0.26
     আমি
    0.25
     hehe
    0.25
     lmao
    0.25
     hahaha
    0.25
     We
    0.24
    POSITIVE LOGITS
     fosters
    0.27
     столь
    0.26
     fostering
    0.25
     comprising
    0.25
    可通过
    0.24
     encompassing
    0.24
    Согласно
    0.24
    有助于
    0.24
     termed
    0.23
     encompasses
    0.23
    Activations Density 0.249%

    No Known Activations