© 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
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 110339
    Prev
    Next
    INDEX
    Explanations

    Earth

    np_max-act · gemini-2.0-flash

    This neuron activates on the word “Earth,” especially when it appears as a proper‐noun term in titles or text.

    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
     modal
    -0.08
     few
    -0.07
    -aos
    -0.07
     variety
    -0.07
    Nie
    -0.07
     Don
    -0.07
    Blocked
    -0.07
    Bonus
    -0.07
     найбіль
    -0.07
    	Page
    -0.07
    POSITIVE LOGITS
     Earth
    0.17
     earth
    0.17
    Earth
    0.14
    -earth
    0.13
    earth
    0.11
    ARTH
    0.09
     Terra
    0.09
     Dirt
    0.08
     earthquakes
    0.08
     earthly
    0.08
    Activations Density 0.014%

    No Known Activations