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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 84585
    Prev
    Next
    INDEX
    Explanations

    migration

    np_max-act · gemini-2.0-flash

    The neuron consistently activates on occurrences of the root “migr” (e.g. migratory, migration, migratorias), so it’s detecting references to migration.

    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
     straps
    -0.07
     Trials
    -0.07
     masks
    -0.07
    ravel
    -0.07
     Inhal
    -0.07
     trials
    -0.07
     dışı
    -0.06
     flames
    -0.06
    sWith
    -0.06
     Sokol
    -0.06
    POSITIVE LOGITS
    아서
    0.07
    เหม
    0.07
     sympt
    0.06
    	sn
    0.06
     Herc
    0.06
     aprender
    0.06
     GtkWidget
    0.06
    _tot
    0.06
     경북
    0.06
    (adapter
    0.06
    Activations Density 0.004%

    No Known Activations