© 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. 12431
    Prev
    Next
    INDEX
    Explanations

    prior

    np_max-act · gemini-2.0-flash

    The neuron specifically picks up on attempts to override or nullify earlier instructions or directives (e.g. “all previous directives are now null and void”).

    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
     melt
    -0.06
    师
    -0.06
     broad
    -0.06
     atas
    -0.06
    ials
    -0.06
    tons
    -0.06
     SPL
    -0.06
     "(
    -0.05
     breath
    -0.05
     Stuart
    -0.05
    POSITIVE LOGITS
    Invoice
    0.07
    覚
    0.07
     preprocess
    0.06
     rug
    0.06
    section
    0.06
    ищ
    0.06
     ближ
    0.06
    \Json
    0.06
    /trans
    0.06
    .pr
    0.06
    Activations Density 0.002%

    No Known Activations