© 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
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 55484
    Prev
    Next
    INDEX
    Explanations

    self-reflection, growth

    np_max-act · gemini-2.0-flash

    The neuron flags directive or instruction verbs (the imperative “Generate,” “Provide,” “Write,” etc.) that signal user commands.

    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
     Penny
    -0.06
     Tower
    -0.06
     oats
    -0.06
    AUTH
    -0.06
     Dialogue
    -0.06
     review
    -0.06
     Robin
    -0.06
    	sum
    -0.06
    �
    -0.06
     guide
    -0.05
    POSITIVE LOGITS
    _NEAR
    0.08
     Ultra
    0.07
    _COMMON
    0.06
     Explosion
    0.06
    _district
    0.06
    -boot
    0.06
    .ext
    0.06
    ทาง
    0.06
     ['./
    0.06
    .localtime
    0.06
    Activations Density 0.040%

    No Known Activations