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. 5-RES-JB
    6. 6764
    Prev
    Next
    INDEX
    Explanations

    phrases related to historical or cultural references regarding buildings or places, especially mentioning the word "vel."

    oai_token-act-pair · gpt-3.5-turbo

    the word "vel," indicating a focus on various meanings or contexts related to "velocity" or "level."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 5-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    5
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Shinra
    -0.70
    olicy
    -0.69
    riad
    -0.69
    BILITIES
    -0.68
     "$:/
    -0.66
     EDITION
    -0.64
    cffffcc
    -0.64
    ribe
    -0.64
    arij
    -0.63
    é¾
    -0.62
    POSITIVE LOGITS
    ocity
    1.22
    vet
    0.86
    edo
    0.79
    icit
    0.79
    istically
    0.79
    vel
    0.77
    theless
    0.76
    lette
    0.71
    ary
    0.71
    arial
    0.70
    Activations Density 0.008%

    No Known Activations