© 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. 7054
    Prev
    Next
    INDEX
    Explanations

    within, in, between

    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
     Directive
    -0.10
    åĽĽçº§
    -0.09
    fin
    -0.09
    (fin
    -0.09
     finances
    -0.09
    forc
    -0.08
    aval
    -0.08
    Directive
    -0.08
     fin
    -0.08
    éĩĿ
    -0.08
    POSITIVE LOGITS
    åĨħå¤ĸ
    0.10
    å¤ļåįĬ
    0.10
    俸
    0.09
    æµ´
    0.09
    insic
    0.09
    æ·±å¤Ħ
    0.09
    rosso
    0.08
    çİ·
    0.08
    çĶŁåij½çļĦ
    0.08
    æĿ¡è¯Ħ论
    0.08
    Activations Density 0.127%

    No Known Activations