© 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 (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 73582
    Prev
    Next
    INDEX
    Explanations

    Code/API interaction

    np_max-act · gemini-2.0-flash

    This neuron activates on the assistant’s internal reasoning markers (e.g. “Thought:”, “Action:”, and similar chain‐of‐thought or tool‐invocation headers).

    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
    arLayout
    -0.07
    }',↵
    -0.07
    ),↵
    -0.07
    Whitespace
    -0.06
    urma
    -0.06
     influx
    -0.06
    }
    ↵
    -0.06
    /y
    -0.06
    eware
    -0.06
    creature
    -0.06
    POSITIVE LOGITS
    E
    0.07
    327
    0.07
     بحث
    0.06
     اس
    0.06
    668
    0.06
     स
    0.06
    Louis
    0.06
    519
    0.06
    NUMBER
    0.06
     nosotros
    0.06
    Activations Density 0.005%

    No Known Activations