© 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. 22873
    Prev
    Next
    INDEX
    Explanations

    rebel

    np_max-act · gemini-2.0-flash

    The neuron detects words and word pieces that express negative personality traits, insults, hostility, or humiliation (e.g., arrogant, cocky, greedy, mean, humiliate).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    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
     Skyrim
    -0.07
    _comment
    -0.07
    вар
    -0.07
    кол
    -0.06
     commenting
    -0.06
    _does
    -0.06
     icing
    -0.06
    _Lean
    -0.06
    	hash
    -0.06
    “These
    -0.06
    POSITIVE LOGITS
    ıyla
    0.06
    xCC
    0.06
     olur
    0.06
     amsterdam
    0.06
    .CON
    0.06
     Bengal
    0.06
     رم
    0.06
     Moto
    0.06
     cantidad
    0.06
    علی
    0.06
    Activations Density 0.140%

    No Known Activations