© 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. 15349
    Prev
    Next
    INDEX
    Explanations

    News articles

    np_max-act · gemini-2.0-flash

    the neuron detects passive-voice auxiliary constructions (forms like "are/were" used in passive phrases such as "are framed," "were found," "have been awarded").

    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
    房企
    -0.07
    combe
    -0.07
    ᆱ
    -0.06
    归属于
    -0.06
    缞
    -0.06
    ライフ
    -0.06
    踌
    -0.06
    最主要的
    -0.06
    녔
    -0.06
    esses
    -0.06
    POSITIVE LOGITS
     Compression
    0.07
    .didReceiveMemoryWarning
    0.07
    learn
    0.07
    (tweet
    0.07
     prosecution
    0.07
     enlightenment
    0.06
    .disconnect
    0.06
    ience
    0.06
    成为中国
    0.06
    =node
    0.06
    Activations Density 0.781%

    No Known Activations