© 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. 55434
    Prev
    Next
    INDEX
    Explanations

    Checking availability

    np_max-act · gemini-2.0-flash

    The neuron is sensitive to code identifiers—variable and function names—highlighting tokens like “unique,” “id,” “while,” and other programming identifiers.

    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
    (ex
    -0.07
    Palindrome
    -0.06
     ("%
    -0.06
     Исп
    -0.06
     contemplating
    -0.06
     BUS
    -0.06
     Orchestra
    -0.06
     //--------------------------------
    -0.06
     CHUNK
    -0.06
     ‪
    -0.06
    POSITIVE LOGITS
     organizace
    0.07
     ori
    0.07
     عص
    0.06
    racak
    0.06
    .EventQueue
    0.06
    ों
    0.06
    pher
    0.06
    .sale
    0.06
     immobil
    0.06
    erah
    0.06
    Activations Density 0.022%

    No Known Activations