© 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. Qwen2.5-7B-IT
    7. Resid Post - 131k
    9. 15-RESID-POST-AA
    11. 27213
    Prev
    Next
    INDEX
    Explanations

    dialogue

    np_max-act · gemini-2.0-flash

    Tokens used as speaker/turn labels or other transcript/metadata markers (speaker tags, "sp", "SPEAKER_", "NAME_", msg/str labels, and similar markup).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-qwen2.5-7b-instruct/resid_post_layer_15/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.15.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     educating
    -0.07
    较快
    -0.07
    .MapFrom
    -0.07
    severity
    -0.07
    HK
    -0.06
    [rand
    -0.06
    _UNSIGNED
    -0.06
     שח
    -0.06
    (choice
    -0.06
    独特
    -0.06
    POSITIVE LOGITS
     assistants
    0.08
    actors
    0.08
    appings
    0.07
     graphs
    0.07
    航行
    0.06
     palette
    0.06
    igs
    0.06
    0.06
    تابع
    0.06
     organisers
    0.06
    Activations Density 0.145%

    No Known Activations