© 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. 15-RESID-POST-AA
    11. 19891
    Prev
    Next
    INDEX
    Explanations

    file references

    np_max-act · gemini-2.0-flash

    This neuron detects conversation/control boundary tokens and role markers (the special <|im_start|>, <|im_end|> and system/user/assistant delimiters).

    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
    _est
    -0.07
    pin
    -0.07
    🔦
    -0.07
    Golden
    -0.07
    wy
    -0.07
    رتبط
    -0.06
    >m
    -0.06
    물을
    -0.06
    points
    -0.06
    headline
    -0.06
    POSITIVE LOGITS
     StatusBar
    0.07
    Maria
    0.07
     Baton
    0.07
     submits
    0.06
     Chatt
    0.06
    🖋
    0.06
    :?
    0.06
    (from
    0.06
     Compilation
    0.06
    .SUCCESS
    0.06
    Activations Density 0.274%

    No Known Activations