© 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. 125274
    Prev
    Next
    INDEX
    Explanations

    stories and characters

    np_max-act · gemini-2.0-flash

    This neuron fires on the storytelling or scenario‐framing language—words that set up or describe the characters’ ongoing experiences (e.g. “allowing,” “experience,” “live,” “characters,” etc.).

    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
     horizontally
    -0.07
    umbn
    -0.07
    senal
    -0.07
     ＼
    -0.06
    CHandle
    -0.06
    ?>
↵
↵
    -0.06
     نمونه
    -0.06
    461
    -0.06
     уд
    -0.06
    /tos
    -0.06
    POSITIVE LOGITS
    kill
    0.07
    们
    0.06
    _css
    0.06
     networks
    0.06
    rais
    0.06
     Networks
    0.06
            
↵        
↵
    0.06
    _health
    0.06
     Crimes
    0.06
     delight
    0.06
    Activations Density 0.002%

    No Known Activations