© 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-3-4B-IT
    5. 17-GEMMASCOPE-2-RES-16K
    7. 12666
    Prev
    Next
    INDEX
    Explanations

    lost in the context of

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-4b-it/resid_post/layer_17_width_16k_l0_medium
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    tained
    2.44
    antiate
    2.33
    ﺎ
    2.32
    ва
    2.31
     তোলে
    2.24
    да
    2.08
    tains
    2.05
     squeezing
    2.05
     faulted
    2.02
    不上
    2.02
    POSITIVE LOGITS
    ти
    2.73
     angeles
    2.70
     osse
    2.64
    ippy
    2.46
    𝐵
    2.45
    inafter
    2.44
     volg
    2.41
     ruim
    2.40
     Gestion
    2.34
     pulgadas
    2.31
    Activations Density 0.045%

    No Known Activations