© 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. 30002
    Prev
    Next
    INDEX
    Explanations

    Names

    np_max-act · gemini-2.0-flash

    It specializes in spotting occurrences of the substring “Gold,” especially in proper names like “Solid Gold.”

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

    capitalized proper names and other named entities (people, bands, model names, titles) across languages.

    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
     von
    -0.06
     midfielder
    -0.06
    }↵↵↵
    -0.06
    ,G
    -0.06
     форму
    -0.06
     who
    -0.06
    ortality
    -0.06
     youngest
    -0.06
     Resistance
    -0.06
     present
    -0.06
    POSITIVE LOGITS
    ेदन
    0.07
     atm
    0.07
     PDO
    0.07
     Qty
    0.06
    目标
    0.06
    اا
    0.06
     名無し
    0.06
    _DSP
    0.06
    (pref
    0.06
    NavLink
    0.06
    Activations Density 0.406%

    No Known Activations