© 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. 10994
    Prev
    Next
    INDEX
    Explanations

    relive, playback, rehearsal

    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
     пÑĢим
    -0.10
    ery
    -0.09
    åħ¼èģĮ
    -0.09
    oper
    -0.09
    plates
    -0.09
    pile
    -0.08
    аÑĪ
    -0.08
    æijŃ
    -0.08
    olid
    -0.08
    衽
    -0.08
    POSITIVE LOGITS
    åĪĥ
    0.09
    ศาส
    0.08
    åĪļæīį
    0.08
     tego
    0.08
    溯
    0.08
    ä»ĬçĶŁ
    0.08
    ti
    0.08
     sincer
    0.08
    å®īçĦ¶
    0.08
    æĦ
    0.08
    Activations Density 0.004%

    No Known Activations