© 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. 110151
    Prev
    Next
    INDEX
    Explanations

    scientific/mathematical text

    np_max-act · gemini-2.0-flash

    The neuron is primarily activated by numerical tokens, especially decimal numbers.

    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
     run
    -0.07
    pom
    -0.07
     mog
    -0.07
    eturn
    -0.06
     Plants
    -0.06
    :`
    -0.06
     thugs
    -0.06
    AVOR
    -0.06
     ding
    -0.06
     offended
    -0.06
    POSITIVE LOGITS
     акту
    0.07
     posicion
    0.06
    альную
    0.06
    اا
    0.06
    semicolon
    0.06
    ‐-
    0.06
    しい
    0.06
    ofday
    0.06
    Informe
    0.06
    /wait
    0.06
    Activations Density 0.095%

    No Known Activations