© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 38501
    Prev
    Next
    INDEX
    Explanations

    Reference materials

    np_max-act · gemini-2.0-flash

    This neuron detects Portuguese phrasing that introduces something “as a reference,” e.g. tokens like “como,” “referência,” and “para” when giving examples or sources.

    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
     효과
    -0.07
    .getComponent
    -0.06
    開発
    -0.06
    ][:
    -0.06
     miraculous
    -0.06
    الق
    -0.06
     compressed
    -0.06
     predecessors
    -0.06
    	format
    -0.06
     collapsed
    -0.06
    POSITIVE LOGITS
    	plt
    0.06
    ,arg
    0.06
     vra
    0.06
    concert
    0.06
    ibili
    0.06
     джер
    0.06
     Tuy
    0.06
     Latinos
    0.06
    isans
    0.06
    porn
    0.06
    Activations Density 0.025%

    No Known Activations