© 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. 30535
    Prev
    Next
    INDEX
    Explanations

    rewards and gains

    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
    è¿Ļä¸ĢçĤ¹
    -0.09
    anni
    -0.08
    æİ¥è¿ŀ
    -0.08
    æļĤåģľ
    -0.08
    Continue
    -0.08
    itti
    -0.08
    ä¸į足以
    -0.08
    ẫ
    -0.08
    çĥĪ士
    -0.08
    srv
    -0.08
    POSITIVE LOGITS
     rewards
    0.13
    æį¢æĿ¥
    0.12
     rewarding
    0.12
    åį´æĺ¯
    0.12
    æĶ¶çĽĬ
    0.11
    åĽŀæĬ¥
    0.11
    reward
    0.11
     reward
    0.11
    Reward
    0.11
     Rewards
    0.11
    Activations Density 0.201%

    No Known Activations