© 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. 109338
    Prev
    Next
    INDEX
    Explanations

    body, health, and diet

    np_max-act · gemini-2.0-flash

    This neuron detects the special control and metadata tokens (e.g. <|start_header_id|>, <|end_header_id|>, <|eot_id|>) used to mark document structure.

    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
    apor
    -0.06
     canh
    -0.06
    (K
    -0.06
    	score
    -0.06
    іту
    -0.06
    basis
    -0.06
     bath
    -0.06
     داد
    -0.06
     mp
    -0.06
    metic
    -0.06
    POSITIVE LOGITS
    廊
    0.06
     shard
    0.06
    JE
    0.06
     trait
    0.06
    ?action
    0.06
    \xc
    0.06
     सम
    0.05
     Xã
    0.05
     perfil
    0.05
    oloj
    0.05
    Activations Density 0.008%

    No Known Activations