© 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. 73811
    Prev
    Next
    INDEX
    Explanations

    expectations

    np_max-act · gemini-2.0-flash

    This neuron detects hedging or speculative language—modal verbs and adverbs expressing uncertainty (e.g., “probably,” “might,” “would”).

    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
    chief
    -0.07
    counter
    -0.07
    processors
    -0.07
    ########
    -0.07
     Suite
    -0.07
    dc
    -0.07
    bras
    -0.06
    ock
    -0.06
    solution
    -0.06
     phối
    -0.06
    POSITIVE LOGITS
    toFixed
    0.07
     віз
    0.06
    スカ
    0.06
     AMC
    0.06
    lacak
    0.06
    .Student
    0.06
     näch
    0.06
     eruption
    0.06
    serrat
    0.06
     won
    0.06
    Activations Density 0.336%

    No Known Activations