© 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. 8922
    Prev
    Next
    INDEX
    Explanations

    defense

    np_max-act · gemini-2.0-flash

    This neuron specifically activates on occurrences of the word “defense” (in any context or casing).

    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
     omit
    -0.07
    遗
    -0.06
     Sweet
    -0.06
    agger
    -0.06
     écrit
    -0.06
     Murray
    -0.06
     obt
    -0.06
    -period
    -0.06
     marking
    -0.06
     Rita
    -0.06
    POSITIVE LOGITS
     defense
    0.14
     defenses
    0.13
     Defense
    0.12
     Defence
    0.11
     defence
    0.10
     defensive
    0.10
    Defense
    0.10
    防
    0.08
    ď
    0.08
     defenders
    0.08
    Activations Density 0.019%

    No Known Activations