© 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. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 28144
    Prev
    Next
    INDEX
    Explanations

    ice melting and thawing

    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
    èģ¿
    -0.10
    OLA
    -0.09
    ivers
    -0.09
    éªĽ
    -0.09
     Concent
    -0.09
    é¢Īæ¤İ
    -0.09
    çĽĸ
    -0.09
    _gid
    -0.08
    åģļå®Į
    -0.08
    è¯ĨåĪ«
    -0.08
    POSITIVE LOGITS
     thaw
    0.10
    èŀįåĮĸ
    0.10
    èŀį
    0.09
    éĶĢæ¯ģ
    0.09
    ç¨Ģ
    0.09
    çħ§æł·
    0.09
    squ
    0.09
    åĪ¶åĨ·
    0.09
    åĨ»ç»ĵ
    0.09
    æľį
    0.08
    Activations Density 0.111%

    No Known Activations