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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 40454
    Prev
    Next
    INDEX
    Explanations

    |

    np_max-act · gemini-2.0-flash

    This neuron detects lines of ASCII‐art or decorative text (e.g. sequences of backslashes, pipes, underscores, etc.) rather than natural language.

    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
     ocur
    -0.07
    قاء
    -0.06
    below
    -0.06
    ROAD
    -0.06
     Salvador
    -0.06
     Partition
    -0.06
     Producto
    -0.06
     Palestinians
    -0.06
    fold
    -0.06
     ocor
    -0.06
    POSITIVE LOGITS
    ८
    0.06
     Wouldn
    0.06
    .Secret
    0.06
    -free
    0.06
    00
    0.06
     battles
    0.06
    (let
    0.06
    ίζει
    0.06
     fontWithName
    0.06
    Những
    0.06
    Activations Density 0.003%

    No Known Activations