Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 86208
    Prev
    Next
    INDEX
    Explanations

    code and math tokens

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on the conjunction “and” when it’s used to link names or list items, especially in a donor list format.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-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
    	Ext
    -0.07
    frontend
    -0.07
    _door
    -0.06
    ุง
    -0.06
    -io
    -0.06
    -twitter
    -0.06
    _PC
    -0.06
    .Item
    -0.06
    ức
    -0.06
     دنی
    -0.06
    POSITIVE LOGITS
     commerc
    0.07
     Mental
    0.07
     Yatırım
    0.07
     AngularFire
    0.07
     unknow
    0.06
     muj
    0.06
     mainAxisAlignment
    0.06
     Viewing
    0.06
     jedná
    0.06
     aj
    0.06
    Activations Density 0.040%

    No Known Activations