© 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. 92036
    Prev
    Next
    INDEX
    Explanations

    dates

    np_max-act · gemini-2.0-flash

    This neuron activates on individual numeric tokens when they appear as parts of dates—especially the digits of year values (e.g. the “2,” “0,” “2,” “3” in “2023”).

    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
     goalt
    -0.07
     stimulation
    -0.06
     smoothly
    -0.06
    Malloc
    -0.06
     biết
    -0.06
     facult
    -0.06
     generator
    -0.06
     useMemo
    -0.06
    (sin
    -0.06
     딸
    -0.06
    POSITIVE LOGITS
     Squad
    0.07
    -record
    0.07
    -tab
    0.07
    social
    0.06
     місто
    0.06
     thuốc
    0.06
    terms
    0.06
    روج
    0.06
    /news
    0.06
     dokument
    0.06
    Activations Density 0.019%

    No Known Activations