© 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. 3816
    Prev
    Next
    INDEX
    Explanations

    justify, justification, rationale

    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
    å´§
    -0.10
    æ®ĵ
    -0.10
    olute
    -0.10
    OLUTE
    -0.09
    æıĽ
    -0.09
    æį¢
    -0.09
    gars
    -0.09
    åŃĺ款
    -0.08
    å¯Łçľĭ
    -0.08
    еÑģÑĤво
    -0.08
    POSITIVE LOGITS
     justification
    0.24
     justify
    0.22
    çIJĨçĶ±
    0.20
    justify
    0.20
    çļĦçIJĨçĶ±
    0.18
     rationale
    0.17
     justified
    0.17
     legitim
    0.13
    æŃ£å½ĵ
    0.13
     excuses
    0.12
    Activations Density 0.053%

    No Known Activations