© 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. 22805
    Prev
    Next
    INDEX
    Explanations

    actually

    np_max-act · gemini-2.0-flash

    instructions that try to “jailbreak” the assistant by forcing a no-limits persona and overriding normal ethical or system constraints.

    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
    可是
    -0.07
     потер
    -0.07
     coorden
    -0.07
    ểm
    -0.07
     True
    -0.06
     hâl
    -0.06
    Highest
    -0.06
    TURE
    -0.06
     extraction
    -0.06
     mesel
    -0.06
    POSITIVE LOGITS
     समय
    0.07
     дев
    0.06
    /authentication
    0.06
     distinguish
    0.06
    ographer
    0.06
    .activities
    0.06
     кислот
    0.06
     ΠΑΝ
    0.06
    .Ab
    0.06
     sinus
    0.06
    Activations Density 0.710%

    No Known Activations