© 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. 62999
    Prev
    Next
    INDEX
    Explanations

    undermine or weaken

    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
    被迫
    -0.09
     Primitive
    -0.09
    andal
    -0.09
    è§ģè¯Ĩ
    -0.09
    æĭį
    -0.09
    èĵĿ天
    -0.08
    _primitive
    -0.08
     primitive
    -0.08
    ials
    -0.08
    强迫
    -0.08
    POSITIVE LOGITS
    åīĬå¼±
    0.16
     undermining
    0.13
     undermined
    0.13
    èħIJèļĢ
    0.12
    çł´åĿı
    0.12
     weakening
    0.12
    æİı
    0.12
     undermine
    0.11
    èĦĨå¼±
    0.11
     undermines
    0.11
    Activations Density 0.109%

    No Known Activations