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

    Neuronpedia

    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. 7484
    Prev
    Next
    INDEX
    Explanations

    Stop words

    np_max-act · gemini-2.0-flash

    The neuron activates on descriptive attribute phrases—especially “known for,” “features,” “provides,” “used for,” etc.—marking clauses that state what something is characterized by or does.

    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
    授
    -0.07
    _med
    -0.07
     tents
    -0.06
     Ok
    -0.06
     Arguments
    -0.06
    Jason
    -0.06
    .DrawString
    -0.06
    :model
    -0.06
     Jason
    -0.06
    .extern
    -0.06
    POSITIVE LOGITS
     solder
    0.07
     JAXBElement
    0.06
    .telegram
    0.06
     foreseeable
    0.06
     万円
    0.06
    	EIF
    0.06
     Visa
    0.06
     �
    0.06
    ihn
    0.05
    bindValue
    0.05
    Activations Density 0.105%

    No Known Activations