© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 11-RESID-POST-AA
    11. 2722
    Prev
    Next
    INDEX
    Explanations

    Addressing another person

    np_max-act · gemini-2.0-flash

    tokens that indicate system-level role or identity statements — i.e., instructions and phrases identifying the model/assistant or its role.

    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_11/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.11.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    installation
    -0.08
    elijke
    -0.08
    idency
    -0.08
     çeşitli
    -0.07
    ผล
    -0.07
    全套
    -0.07
    erable
    -0.07
    ophilia
    -0.07
    知識
    -0.07
    的企业
    -0.07
    POSITIVE LOGITS
     curb
    0.07
     Dzi
    0.07
    便宜
    0.07
    -kit
    0.06
    	sig
    0.06
    .rs
    0.06
    urban
    0.06
     Await
    0.06
     Bình
    0.06
     thơ
    0.06
    Activations Density 2.891%

    No Known Activations