© 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 (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 68910
    Prev
    Next
    INDEX
    Explanations

    meals

    np_max-act · gemini-2.0-flash

    The neuron activates on tokens referring to clock times and meals (e.g., “12:00 PM,” “lunch,” “dinner”), i.e. time-of-day markers associated with eating.

    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
     материал
    -0.07
    您
    -0.07
     Род
    -0.06
    .easing
    -0.06
    Elf
    -0.06
    OrderId
    -0.06
    Ip
    -0.06
    ByName
    -0.06
     Clips
    -0.06
     Π
    -0.06
    POSITIVE LOGITS
    	ch
    0.07
     abolished
    0.07
     Hö
    0.06
    ,function
    0.06
    	function
    0.06
     sonrasında
    0.06
    PATCH
    0.06
    loid
    0.06
     function
    0.06
     هزینه
    0.06
    Activations Density 0.011%

    No Known Activations