© 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. DeepSeek-R1-Dist-Llama-8B
    5. 21-LLAMASCOPE-SLIMPJ-OPENR1-RES-32K
    7. 29721
    Prev
    Next
    INDEX
    Explanations

    commas near numbers or mathematical references, possibly indicating complex or combined mathematical steps

    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
    atten
    -0.07
    iaux
    -0.07
     eskort
    -0.07
    awner
    -0.07
    addin
    -0.06
    awi
    -0.06
    à¸Ľà¸£
    -0.06
    åĶ¯
    -0.06
     nors
    -0.06
    vell
    -0.06
    POSITIVE LOGITS
     or
    0.15
     but
    0.14
     maybe
    0.11
     hence
    0.10
     since
    0.10
     although
    0.09
     if
    0.09
     perhaps
    0.09
     like
    0.09
     which
    0.09
    Activations Density 0.297%

    No Known Activations