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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 40986
    Prev
    Next
    INDEX
    Explanations

    mindfulness and consideration

    np_max-act · gemini-2.0-flash

    The neuron fires on polite disclaimers and hedging language (e.g. “It’s important to remember,” “be mindful,” “I’m sorry”).

    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
    цями
    -0.07
     MF
    -0.07
     PVC
    -0.06
     Profession
    -0.06
     raided
    -0.06
    _epsilon
    -0.06
     coffin
    -0.06
     attravers
    -0.06
     rupt
    -0.06
     Lee
    -0.06
    POSITIVE LOGITS
    baar
    0.07
    -local
    0.07
    getic
    0.07
    .opend
    0.06
    .Drawing
    0.06
     thoughtful
    0.06
    ikon
    0.06
    ент
    0.06
    ulate
    0.06
    	q
    0.06
    Activations Density 0.015%

    No Known Activations