© 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. 126724
    Prev
    Next
    INDEX
    Explanations

    conversational exchanges

    np_max-act · gemini-2.0-flash

    The neuron fires on tokens at the start of user questions—i.e. it detects when the user is asking a question.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    whether a sentence is an interrogative (i.e., marks or occurs at the start of a question).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    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
    方法
    -0.07
     CE
    -0.07
    airport
    -0.07
    注意
    -0.06
    YYYY
    -0.06
     heiß
    -0.06
    _cert
    -0.06
     Abdul
    -0.06
    험
    -0.06
    _MULTI
    -0.06
    POSITIVE LOGITS
    latitude
    0.06
    Suit
    0.06
    .emplace
    0.06
    .banner
    0.06
    Justice
    0.06
     ба
    0.06
     */
↵
↵
    0.06
     нов
    0.06
     заяви
    0.06
    _false
    0.06
    Activations Density 0.068%

    No Known Activations