© 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. 58055
    Prev
    Next
    INDEX
    Explanations

    acquitted or cleared

    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
    åĬ¡
    -0.09
    åīįæīĢæľªæľī
    -0.09
    åĪĽ
    -0.09
     jelly
    -0.09
     Kre
    -0.08
    æ¡ĮåŃIJ
    -0.08
    泪水
    -0.08
    æĤĦ
    -0.08
    缮
    -0.08
    toPromise
    -0.08
    POSITIVE LOGITS
    åħįè´£
    0.15
    解èĦ±
    0.15
    æĹłè¾ľ
    0.14
     innocence
    0.13
    åħį
    0.12
    éĢįéģ¥
    0.11
     innocent
    0.11
     blame
    0.11
     acquitted
    0.11
    éĢĥèĦ±
    0.11
    Activations Density 0.073%

    No Known Activations