© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 59390
    Prev
    Next
    INDEX
    Explanations

    began/started

    np_max-act · gemini-2.0-flash

    information about journalists and their professional backgrounds.

    oai_token-act-pair · gpt-4o-mini
    Triggered by @xinyanhu8

    The neuron detects terms marking career milestones and timeline transitions—verbs and phrases like “appointed,” “started,” “began,” “prior to,” and “joining” that signal someone’s job or career progression.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
    urdu
    -0.07
    Drag
    -0.07
     Dan
    -0.06
    ictim
    -0.06
    Як
    -0.06
     Dome
    -0.06
    ideshow
    -0.06
     Italy
    -0.06
     drain
    -0.06
    tx
    -0.06
    POSITIVE LOGITS
    ]=>
    0.07
     çat
    0.07
    <Real
    0.07
    unsqueeze
    0.06
     اص
    0.06
    势
    0.06
    跨
    0.06
     cas
    0.06
    ]}>↵
    0.06
    _UTIL
    0.06
    Activations Density 0.033%

    No Known Activations