© 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. Qwen3.5-4B
    5. 15-RES-MATRYOSHKA-65K
    7. 21817
    Prev
    Next
    INDEX
    Explanations

    the question is

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    decoderesearch/qwen-3.5-saes/qwen-3.5-4b
    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
    elis
    -0.06
    线æĢ§
    -0.05
    ä¼ij
    -0.05
     mell
    -0.05
    merksam
    -0.05
    ubes
    -0.05
    ation
    -0.05
    achu
    -0.05
    å¹ħ
    -0.05
     mo
    -0.05
    POSITIVE LOGITS
    --
    0.06
    alternative
    0.06
    den
    0.06
     betr
    0.06
     alternative
    0.06
    view
    0.05
     reporta
    0.05
    say
    0.05
     suppose
    0.05
    env
    0.05
    Activations Density 0.006%

    No Known Activations