© 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. 34-GEMMASCOPE-RES-131K
    7. 31832
    Prev
    Next
    INDEX
    Explanations

    friendship and connections

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

    The neuron responds to affix-like fragments and hyphens in words—e.g. common prefixes (“In-”), suffixes (“-able”) and word-internal hyphens.

    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
     AppCompatTheme
    -1.17
    问题
    -1.13
    他们
    -1.09
    halle
    -1.03
     mentality
    -1.03
    体验
    -1.03
    Disclaimer
    -1.03
    genöss
    -1.02
    این
    -1.00
     épis
    -0.99
    POSITIVE LOGITS
     many
    1.30
     your
    1.27
     vždy
    1.24
     our
    1.20
     will
    1.15
     often
    1.14
    Both
    1.13
     and
    1.12
    Сю
    1.11
     both
    1.10
    Activations Density 0.007%

    No Known Activations