© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 36360
    Prev
    Next
    INDEX
    Explanations

    fire and construction

    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
    åĶ¾
    -0.10
    byss
    -0.09
     pair
    -0.09
    Ñŀ
    -0.08
    å¿ĥæĥħ
    -0.08
    ltk
    -0.08
    _inline
    -0.08
    çł£
    -0.08
    pressão
    -0.08
    åı£å¾Ħ
    -0.08
    POSITIVE LOGITS
    çģ«çģ¾
    0.19
    çĩĥçĥ§
    0.19
    大çģ«
    0.19
    çģŃçģ«
    0.18
    æ¶Īéĺ²
    0.18
    æ¶Īéĺ²å®īåħ¨
    0.17
     arson
    0.17
     Fire
    0.16
     fire
    0.16
    Fire
    0.16
    Activations Density 0.153%

    No Known Activations