© 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. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Qwen2.5-7B-IT
    7. Resid Post - 131k
    9. 15-RESID-POST-AA
    11. 5194
    Prev
    Next
    INDEX
    Explanations

    Tips and tricks

    np_max-act · gemini-2.0-flash

    Tokens that mark numbered list items or step-by-step instruction openings (e.g., numerals with punctuation and the initial words of procedural tips).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-qwen2.5-7b-instruct/resid_post_layer_15/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.15.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     sleeping
    -0.07
    共鸣
    -0.07
    .lb
    -0.06
     البيت
    -0.06
    床上
    -0.06
     Lingu
    -0.06
     spiked
    -0.06
    天猫
    -0.06
    דלת
    -0.06
     Parm
    -0.06
    POSITIVE LOGITS
    ớ
    0.08
    .optimize
    0.08
    �
    0.07
    출
    0.07
    さ
    0.07
    езультат
    0.07
     ما
    0.07
     değerl
    0.07
    	M
    0.07
    monto
    0.07
    Activations Density 0.071%

    No Known Activations