© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 89254
    Prev
    Next
    INDEX
    Explanations

    Online articles

    np_max-act · gemini-2.0-flash

    This neuron detects the special control/boundary tokens that mark the end of a user header and the start of the assistant’s reply.

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

    the start of an assistant-generated reply (activation at assistant/response boundary tokens).

    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
     Higgins
    -0.07
    -success
    -0.06
     기록
    -0.06
     дослід
    -0.06
    šov
    -0.06
    _predictions
    -0.06
     Ок
    -0.06
    损
    -0.06
    erken
    -0.06
     OnInit
    -0.06
    POSITIVE LOGITS
     тиж
    0.07
    )?↵
    0.07
    ?↵
    0.07
    -automatic
    0.07
            
↵
    0.07
     amatør
    0.06
    ilated
    0.06
     realities
    0.06
    ธาน
    0.06
    ifference
    0.06
    Activations Density 0.079%

    No Known Activations