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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 319
    Prev
    Next
    INDEX
    Explanations

    Directions

    np_max-act · gemini-2.0-flash

    This neuron detects mentions of cardinal directions (e.g. north, south, east, west) used in describing geographic boundaries.

    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
    tridge
    -0.07
    365
    -0.06
    	           
    -0.06
    319
    -0.06
     رض
    -0.06
    'an
    -0.06
    767
    -0.06
    ủ
    -0.06
    worth
    -0.06
    úa
    -0.06
    POSITIVE LOGITS
    ータ
    0.07
    .Ok
    0.07
     Macron
    0.07
     사건
    0.07
     trek
    0.07
    	T
    0.07
    County
    0.06
     borrower
    0.06
    =t
    0.06
     [+
    0.06
    Activations Density 0.017%

    No Known Activations