© 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. 126761
    Prev
    Next
    INDEX
    Explanations

    two or three

    np_max-act · gemini-2.0-flash

    The neuron responds to the words “two,” “three,” and “four” when they specify how many letters are being picked.

    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
    	HX
    -0.07
    وقت
    -0.07
    .FormattingEnabled
    -0.06
     تنها
    -0.06
    (HTTP
    -0.06
     ○
    -0.06
    emás
    -0.06
    での
    -0.06
     restau
    -0.06
     meals
    -0.06
    POSITIVE LOGITS
    .getString
    0.08
    ARTH
    0.08
    seo
    0.07
    нак
    0.07
    арь
    0.07
    าผ
    0.07
     pain
    0.07
    ModelCreating
    0.07
    launch
    0.06
    .getFloat
    0.06
    Activations Density 0.001%

    No Known Activations