© 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
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 126808
    Prev
    Next
    INDEX
    Explanations

    requests and demands

    np_max-act · gemini-2.0-flash

    The neuron activates on foreign‐language (non-English) word fragments—particularly Slavic/Slovene tokens with diacritics.

    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
     Lux
    -0.06
     Ner
    -0.06
     Applied
    -0.06
     grinder
    -0.06
    _rd
    -0.06
     election
    -0.06
     considering
    -0.06
     applied
    -0.06
     possibility
    -0.06
    ita
    -0.06
    POSITIVE LOGITS
    ději
    0.07
    /css
    0.07
    acomment
    0.07
    navbar
    0.06
    taient
    0.06
     dk
    0.06
     ¡
    0.06
    .sap
    0.06
    _kategori
    0.06
    (iv
    0.06
    Activations Density 0.177%

    No Known Activations