© 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
    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. 42571
    Prev
    Next
    INDEX
    Explanations

    Expressing opinions

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on the function‐word pattern of a “not … but …” contrastive construction (e.g. “is not to … but to …”).

    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
    Serializer
    -0.07
    subtotal
    -0.06
     succes
    -0.06
     friday
    -0.06
    !");↵
    -0.06
    _middle
    -0.06
    middle
    -0.06
    \"><
    -0.06
    반
    -0.06
    (Resource
    -0.06
    POSITIVE LOGITS
    áo
    0.07
     occup
    0.06
    िए
    0.06
    осред
    0.06
    개의
    0.06
    okane
    0.06
    itchen
    0.06
    setContent
    0.06
     rarely
    0.06
     สล
    0.06
    Activations Density 0.054%

    No Known Activations