© 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. 7758
    Prev
    Next
    INDEX
    Explanations

    variable types

    np_max-act · gemini-2.0-flash

    The neuron activates on type annotations—especially the keyword “string” (and other return‐type tokens) in function signatures or declarations.

    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
    ταση
    -0.07
     LDS
    -0.07
    sburg
    -0.06
     खर
    -0.06
    _CODES
    -0.06
    kommen
    -0.06
     productId
    -0.06
    ,"
    -0.06
     haired
    -0.06
    CU
    -0.06
    POSITIVE LOGITS
    ]<=
    0.06
    Refs
    0.06
     تز
    0.06
     influencing
    0.06
     Olive
    0.06
    :length
    0.06
    :Boolean
    0.06
     confessed
    0.06
    	restore
    0.06
    boa
    0.06
    Activations Density 0.018%

    No Known Activations