Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 126635
    Prev
    Next
    INDEX
    Explanations

    News and technical articles

    np_max-act · gemini-2.0-flash

    This neuron activates on the “assistant” role indicator token in the conversation metadata.

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

    the assistant speaker/header tag in chat-style metadata indicating assistant responses.

    oai_token-act-pair · gpt-5Triggered 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
    하여
    -0.07
    渐
    -0.07
    -containing
    -0.07
    	Array
    -0.07
    řen
    -0.06
    По
    -0.06
     изуч
    -0.06
     상
    -0.06
    لس
    -0.06
    ンフ
    -0.06
    POSITIVE LOGITS
    iale
    0.07
     موضوع
    0.06
     calmly
    0.06
    coma
    0.06
    .impl
    0.06
     Proposal
    0.06
     proposal
    0.06
     Vietnam
    0.06
    ator
    0.06
    createTime
    0.06
    Activations Density 0.017%

    No Known Activations