© 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. 44140
    Prev
    Next
    INDEX
    Explanations

    act like

    np_max-act · gemini-2.0-flash

    This neuron activates on tokens in jailbreak-style instructions telling the model to “act as” or “simulate” a particular mode (e.g., Developer/DAN mode).

    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
     deutschland
    -0.07
    равиль
    -0.06
     захоп
    -0.06
     sweeps
    -0.06
     zum
    -0.06
     abolish
    -0.06
    _gray
    -0.06
     insurer
    -0.06
     있도록
    -0.06
     bilingual
    -0.06
    POSITIVE LOGITS
     khuôn
    0.07
    дат
    0.07
    YG
    0.06
    دث
    0.06
    hg
    0.06
    _rate
    0.06
    ンズ
    0.06
    heed
    0.06
     CX
    0.06
    _CLASSES
    0.06
    Activations Density 0.012%

    No Known Activations