© 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. 32949
    Prev
    Next
    INDEX
    Explanations

    fire and building materials

    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
    çĨĦ
    -0.12
    coal
    -0.11
     coal
    -0.11
    çģŃ
    -0.10
    çĶŁçĶŁ
    -0.10
    çĩĥçĥ§
    -0.10
    çģ«çĦ°
    -0.10
    ä¸ĢåįĬ
    -0.10
    çĥ
    -0.10
    çħ¤
    -0.09
    POSITIVE LOGITS
    éĽģ
    0.09
    轸
    0.09
    èµĦè´¨
    0.09
    ä½IJ
    0.09
    å°¾
    0.09
     tails
    0.08
    ear
    0.08
    æ´Ĺ
    0.08
    TAIL
    0.08
    lip
    0.08
    Activations Density 0.025%

    No Known Activations