© 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. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 51084
    Prev
    Next
    INDEX
    Explanations

    sweetness or sweet flavors

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    adamkarvonen/qwen3-32b-saes/saes_Qwen_Qwen3-32B_batch_top_k/resid_post_layer_32
    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
    unch
    -0.09
    ìĤ¬íķŃ
    -0.09
    Ľå»º
    -0.09
    UGE
    -0.09
    å¾®éĩıåħĥç´ł
    -0.09
    itional
    -0.09
    undi
    -0.08
    aid
    -0.08
    igned
    -0.08
    baugh
    -0.08
    POSITIVE LOGITS
    ening
    0.19
    ened
    0.18
    eners
    0.13
    ness
    0.13
    NESS
    0.11
    stakes
    0.11
    ener
    0.11
    (er
    0.11
    fony
    0.10
    ly
    0.10
    Activations Density 0.008%

    No Known Activations