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

    Neuronpedia

    APISteerSAE EvalsBlog/PodcastNEWSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlog/PodcastGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 21090
    Prev
    Next
    INDEX
    Explanations

    instances where something or someone is currently in a state of being or doing something

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "being."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    rote
    -0.70
    akh
    -0.62
    chrome
    -0.60
    ruce
    -0.59
    arta
    -0.59
    tex
    -0.59
    uristic
    -0.59
    SELECT
    -0.59
    rone
    -0.57
    dds
    -0.57
    POSITIVE LOGITS
     being
    3.13
    being
    2.63
     Being
    2.14
    Being
    1.88
     having
    1.79
     becoming
    1.77
     appearing
    1.55
    having
    1.51
     needing
    1.49
     undergoing
    1.39
    Activations Density 0.073%

    No Known Activations