© 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. 32672
    Prev
    Next
    INDEX
    Explanations

    comprises a

    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
    çīĮåŃIJ
    -0.11
    åľ®
    -0.09
    éű
    -0.09
    oce
    -0.09
    ropri
    -0.09
    ĺ
    -0.09
    ró
    -0.09
    راء
    -0.08
    rafted
    -0.08
    é¹Ń
    -0.08
    POSITIVE LOGITS
    forth
    0.13
    cial
    0.09
    nable
    0.09
    sic
    0.08
    çĵł
    0.08
    antas
    0.08
    ure
    0.08
    å·±
    0.08
    avel
    0.08
    allo
    0.08
    Activations Density 0.024%

    No Known Activations