© 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. 55935
    Prev
    Next
    INDEX
    Explanations

    Course descriptions

    np_max-act · gemini-2.0-flash

    This neuron detects learner-focused outcome language in course descriptions—words like “you,” “will,” “to,” “able,” and “become” that frame program benefits or objectives.

    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
    igrate
    -0.07
     )
    -0.07
     которая
    -0.07
    σμό
    -0.07
    uste
    -0.07
    TEGR
    -0.06
     byte
    -0.06
     их
    -0.06
     Firewall
    -0.06
     discourse
    -0.06
    POSITIVE LOGITS
     jugg
    0.07
     працю
    0.07
     seaw
    0.06
     แพ
    0.06
     φορ
    0.06
    ▋
    0.06
     chờ
    0.06
    (cm
    0.06
     enam
    0.06
    站在
    0.06
    Activations Density 0.030%

    No Known Activations