© 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. Qwen2.5-7B-IT
    5. 20-MATRYOSHKA-65K
    7. 3334
    Prev
    Next
    INDEX
    Explanations

    cut

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    chanind/qwen2.5-7B-it-layer-20-saes/pile/matryoshka/k-100
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Oversight
    -0.24
     RedirectTo
    -0.23
     suç
    -0.22
    forgettable
    -0.22
     verschied
    -0.22
     florida
    -0.22
    Republicans
    -0.22
    Hôtel
    -0.21
     erotica
    -0.21
     kald
    -0.21
    POSITIVE LOGITS
    åī²
    0.36
     cut
    0.34
     cuts
    0.33
    ä¸ĢåĪĢ
    0.33
    åĪĢ
    0.33
     cutting
    0.33
    cut
    0.33
    cuts
    0.33
    åĪĩ
    0.33
    åĪĩåī²
    0.33
    Activations Density 0.441%

    No Known Activations