© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 73801
    Prev
    Next
    INDEX
    Explanations

    sexual

    np_max-act · gemini-2.0-flash

    The neuron activates on words that label or describe NSFW content—e.g. terms denoting sexual explicitness, adult material, violence, or other explicit content.

    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
     customer
    -0.07
     emin
    -0.06
     sevent
    -0.06
    켓
    -0.06
    iterals
    -0.06
    customerId
    -0.06
    customer
    -0.06
    dato
    -0.06
    पत
    -0.06
     archae
    -0.06
    POSITIVE LOGITS
     intelligent
    0.07
    	stop
    0.06
    .High
    0.06
    PROGRAM
    0.06
     πρό
    0.06
    	transform
    0.06
     กรกฎาคม
    0.06
    ạy
    0.06
     时间
    0.06
    (y
    0.06
    Activations Density 0.022%

    No Known Activations