© 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. 63329
    Prev
    Next
    INDEX
    Explanations

    been followed by adverbs

    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
    .Act
    -0.10
    çľĭå¾ħ
    -0.09
    DOM
    -0.09
    ibble
    -0.09
    Ĥæķ°
    -0.09
    entially
    -0.09
    ä½ľä¸ļ
    -0.09
    è´¡çĮ®
    -0.09
    tober
    -0.09
    eled
    -0.09
    POSITIVE LOGITS
     further
    0.11
    è¿Ľä¸ĢæŃ¥
    0.11
    移åĪ°
    0.09
     Gad
    0.08
    转移åĪ°
    0.08
    å¿ĥæĦ¿
    0.08
     Graves
    0.08
    æī©å¤§
    0.08
     followed
    0.08
     dev
    0.08
    Activations Density 0.240%

    No Known Activations