Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 93082
    Prev
    Next
    INDEX
    Explanations

    you your

    np_max-act · gemini-2.0-flash

    This neuron detects structural or metadata tokens introduced by the system or conversation instructions (e.g. role-play directives and header/footer markers).

    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
    申请
    -0.07
     Voter
    -0.07
    ocols
    -0.07
    ž
    -0.06
    วร
    -0.06
     tex
    -0.06
     achievable
    -0.06
     viel
    -0.06
     (<
    -0.06
     radial
    -0.06
    POSITIVE LOGITS
     نوفمبر
    0.06
     магаз
    0.06
    ,node
    0.06
    她的
    0.06
    reet
    0.06
     листоп
    0.06
     кожи
    0.05
    利用
    0.05
     machines
    0.05
     Restore
    0.05
    Activations Density 0.053%

    No Known Activations