© 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. Qwen3.5-9B Base
    5. 15-QWENSCOPE-RES-64K
    7. 26649
    Prev
    Next
    INDEX
    Explanations

    would, could, would be

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    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
    erior
    -0.09
    izzano
    -0.09
    ooter
    -0.09
    á»±u
    -0.08
    pery
    -0.08
    æīĢ以å½ĵ
    -0.08
    upos
    -0.08
    adox
    -0.08
    æĪij们å°Ĩåľ¨
    -0.08
    æĹ¶èµ·
    -0.08
    POSITIVE LOGITS
     wouldn
    0.12
     would
    0.12
     Would
    0.11
     could
    0.11
    would
    0.11
     wäre
    0.11
    Would
    0.11
     Could
    0.10
    Could
    0.10
    could
    0.10
    Activations Density 0.329%

    No Known Activations