© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 56092
    Prev
    Next
    INDEX
    Explanations

    transition

    np_max-act · gemini-2.0-flash

    references to implementation details in coding or programming contexts.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    The neuron fires on temporal or conditional discourse markers and their nearby action words—especially “as soon as,” “when,” “after,” etc.—that introduce a new clause or event.

    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
    ushort
    -0.07
     selectable
    -0.07
    권
    -0.07
     §
    -0.06
    "fmt
    -0.06
     Kami
    -0.06
     osoby
    -0.06
    टन
    -0.06
     zaten
    -0.06
    nych
    -0.06
    POSITIVE LOGITS
    ัตว
    0.07
     lining
    0.06
     differed
    0.06
    チ
    0.06
     Vir
    0.06
    İT
    0.06
     childs
    0.06
     Глав
    0.06
    egrate
    0.05
    IK
    0.05
    Activations Density 0.035%

    No Known Activations