© 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. 803
    Prev
    Next
    INDEX
    Explanations

    people

    np_max-act · gemini-2.0-flash

    discussions about the emotional impact of stress and personal experiences.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    The neuron detects vocative or audience-address terms such as “friends,” “people,” and “folks.”

    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
    /Resources
    -0.08
    Repository
    -0.07
    hosts
    -0.06
     Sylv
    -0.06
    Tumblr
    -0.06
    šek
    -0.06
    Instr
    -0.06
    xFFFF
    -0.06
     André
    -0.06
     synthesis
    -0.06
    POSITIVE LOGITS
    ैच
    0.08
     vv
    0.07
    [frame
    0.07
     temps
    0.06
    (=)
    0.06
     Sexe
    0.06
    -gray
    0.06
     previously
    0.06
    soon
    0.06
     baby
    0.06
    Activations Density 0.023%

    No Known Activations