© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. DeepSeek-R1-Dist-Llama-8B
    5. 21-LLAMASCOPE-SLIMPJ-OPENR1-RES-32K
    7. 27570
    Prev
    Next
    INDEX
    Explanations

    It's difficult to find a common thread, but it seems to activate for sentences where the speaker is reasoning something out, and/or telling some kind of story

    oai_token-act-pair · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    fnlp/Llama-Scope-R1-Distill/400M-Slimpajama-400M-OpenR1-Math-220k/L21R
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Hzfinfdu/SlimPajama-3B and open-r1/OpenR1-Math-220k
    Features
    32,768
    Data Type
    float32
    Hook Name
    blocks.21.hook_resid_post
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    abant
    -0.06
    wers
    -0.06
    agus
    -0.06
     kys
    -0.06
    IME
    -0.06
     '..',
    -0.06
    ाà¤ĸ
    -0.06
    /cop
    -0.06
    ledge
    -0.06
    /generated
    -0.06
    POSITIVE LOGITS
     itself
    0.42
     alone
    0.28
     themselves
    0.28
    alone
    0.22
     herself
    0.21
     Alone
    0.21
    èĩªèº«
    0.19
     himself
    0.18
     unto
    0.17
    -alone
    0.16
    Activations Density 0.051%

    No Known Activations