Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 1536
    Prev
    Next
    INDEX
    Explanations

    news articles

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on the key reporting‐verb phrases in news lead sentences—words like “statement,” “appeal,” or “dealing” together with the following “to + verb” infinitive that introduce the article’s main action.

    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
    ]:↵
    -0.07
    едж
    -0.07
    �
    -0.07
     ورد
    -0.06
     NAS
    -0.06
     برنامه
    -0.06
    		    	
    -0.06
     Blonde
    -0.06
     OTP
    -0.06
    �
    -0.06
    POSITIVE LOGITS
    车
    0.07
    援
    0.07
     gọn
    0.06
     phê
    0.06
    (named
    0.06
     bringing
    0.06
     serg
    0.06
     enamel
    0.06
    _BOX
    0.06
    .Tool
    0.06
    Activations Density 0.084%

    No Known Activations