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. 8268
    Prev
    Next
    INDEX
    Explanations

    to

    np_max-act · gemini-2.0-flash

    The neuron detects tokens in expressions of concern—especially the word “worry” (and its immediate context).

    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
    (progress
    -0.07
    -method
    -0.07
    алом
    -0.07
     Fame
    -0.06
    ман
    -0.06
     οικο
    -0.06
     Ramsey
    -0.06
     fluor
    -0.06
     Woods
    -0.06
    リ
    -0.06
    POSITIVE LOGITS
     челов
    0.07
    UGHT
    0.06
    	ref
    0.06
     �
    0.06
     muss
    0.06
    fortawesome
    0.06
     нельзя
    0.06
    Codes
    0.06
    Convert
    0.06
    Thought
    0.06
    Activations Density 0.014%

    No Known Activations