© 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
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 129829
    Prev
    Next
    INDEX
    Explanations

    signs

    np_max-act · gemini-2.0-flash

    This neuron activates on phrases that signal an indication (e.g. “sign(s) that … is/are …”)—in other words, it spots language warning “sign that…” or “signs that…” followed by “is” plus a verb.

    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
    					
    ↵
    -0.06
     ){
    ↵
    -0.06
    .ops
    -0.06
     nurses
    -0.06
    '=>$
    -0.06
     світу
    -0.06
    igs
    -0.06
    ادث
    -0.06
     talked
    -0.06
    ità
    -0.06
    POSITIVE LOGITS
     tritur
    0.07
     Triumph
    0.06
     Tup
    0.06
     Silent
    0.06
    Authentication
    0.06
     rocking
    0.06
     bootloader
    0.06
     Live
    0.06
    %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
    0.06
    _curr
    0.06
    Activations Density 0.002%

    No Known Activations