© 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. 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. 63548
    Prev
    Next
    INDEX
    Explanations

    A strong signal for conversation/role boundary tokens — it fires on the special markers that start or end user/assistant turns (the <|im_start|>/<|im_end|>-style tokens).

    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
     Hemp
    -0.08
    玤
    -0.06
    욧
    -0.06
    {}{↵
    -0.06
     Boh
    -0.06
    _statistics
    -0.06
    .commit
    -0.06
     Farm
    -0.06
    �
    -0.06
    /high
    -0.06
    POSITIVE LOGITS
     noticias
    0.08
     Tehran
    0.07
     |\
    0.07
    quences
    0.07
    oop
    0.07
     Narc
    0.06
    (ic
    0.06
     setContent
    0.06
    _AC
    0.06
    GC
    0.06
    Activations Density 0.069%

    No Known Activations