© 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. 17264
    Prev
    Next
    INDEX
    Explanations

    elevator or oval or atholic

    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
    æłħ
    -0.10
    æijĨåľ¨
    -0.09
    lical
    -0.08
    æ³¼
    -0.08
    isman
    -0.08
    opot
    -0.08
    å°Ĭ严
    -0.08
    _backward
    -0.08
    鬣
    -0.08
    CELL
    -0.08
    POSITIVE LOGITS
    çĶµæ¢¯
    0.26
     elevator
    0.22
    楼å±Ĥ
    0.15
    轿
    0.15
     Elev
    0.14
    levator
    0.12
     elev
    0.12
    楼å®ĩ
    0.12
    大楼
    0.11
    åįĩéĻį
    0.11
    Activations Density 0.034%

    No Known Activations