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. 215
    Prev
    Next
    INDEX
    Explanations

    the

    np_max-act · gemini-2.0-flash

    legal or contractual terms related to agreements and policies.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    The neuron activates on legal‐agreement language—mentions of “Terms of Service,” “Privacy Policy,” or similar consent/terms‐of‐use phrases.

    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
     incorporating
    -0.06
    項目
    -0.06
     Types
    -0.06
    ئيس
    -0.06
     mercenaries
    -0.06
     emergencies
    -0.06
     yours
    -0.06
    Ideal
    -0.06
     anticipating
    -0.06
     cramped
    -0.06
    POSITIVE LOGITS
    ίνει
    0.07
     Sé
    0.07
     Jul
    0.07
     curled
    0.06
     छ
    0.06
    Jean
    0.06
    	cout
    0.06
    周
    0.06
    %</
    0.06
     Linh
    0.06
    Activations Density 0.006%

    No Known Activations