© 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. 21890
    Prev
    Next
    INDEX
    Explanations

    grape or fruit

    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
    ally
    -0.10
    ateur
    -0.10
    ãĤĥ
    -0.10
    çĪ±å¿ĥ
    -0.09
    é©»æĿij
    -0.08
    ders
    -0.08
    åķ¡
    -0.08
    sex
    -0.08
    ial
    -0.08
     rud
    -0.08
    POSITIVE LOGITS
    vine
    0.13
    fruit
    0.10
    ENER
    0.10
    cake
    0.10
    alchemy
    0.09
    lets
    0.09
    å¿Į
    0.09
    ива
    0.09
    iever
    0.09
    åľ¨è¿ĩæ¸¡æľŁ
    0.09
    Activations Density 0.008%

    No Known Activations