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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. DeepSeek-R1-Dist-Llama-8B
    3. 15-LLAMASCOPE-SLIMPJ-OPENR1-RES-32K
    4. 777
    Prev
    Next
    INDEX
    Explanations

    C-style programming keywords and code-related terminology.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @trainjumperx

    terms related to mathematical concepts and functions

    oai_token-act-pair · gpt-4o-mini
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    fnlp/Llama-Scope-R1-Distill/400M-Slimpajama-400M-OpenR1-Math-220k/L15R
    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.15.hook_resid_post
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    _salt
    -0.06
    cano
    -0.06
    akes
    -0.06
    ĩa
    -0.06
     xsi
    -0.06
    rint
    -0.06
     dane
    -0.06
    âī¡
    -0.06
     нÑĸж
    -0.06
    olid
    -0.06
    POSITIVE LOGITS
    /non
    0.07
     leve
    0.07
    BILE
    0.07
    elly
    0.07
    ector
    0.06
    ercul
    0.06
    itz
    0.06
    =true
    0.06
     trip
    0.06
    .opacity
    0.06
    Activations Density 0.217%

    No Known Activations