© 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. 42710
    Prev
    Next
    INDEX
    Explanations

    punctuation

    np_max-act · gemini-2.0-flash

    This neuron detects structural/chat‐format tokens (e.g. meta markers like “<|end_header_id|>,” speaker tags, and other non–content tokens).

    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
    modation
    -0.07
     analysed
    -0.07
     Teeth
    -0.06
     fierc
    -0.06
    iatrics
    -0.06
     contemporary
    -0.06
    noon
    -0.06
     fiberglass
    -0.06
    ザ
    -0.06
     divorced
    -0.05
    POSITIVE LOGITS
    _Dec
    0.08
     โดย
    0.07
     chết
    0.07
     إذ
    0.07
    -commit
    0.06
    -Origin
    0.06
     jean
    0.06
    _se
    0.06
     Δι
    0.06
    _entropy
    0.06
    Activations Density 0.004%

    No Known Activations