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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 3107
    Prev
    Next
    INDEX
    Explanations

    words related to renewal or continued engagement with an activity or commitment

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "renew" and related concepts or actions

    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
    lined
    -0.78
    nosis
    -0.73
    vantage
    -0.73
    visors
    -0.68
    sembly
    -0.68
     Militia
    -0.68
    athlon
    -0.67
    eker
    -0.67
    hetto
    -0.66
     Gauntlet
    -0.66
    POSITIVE LOGITS
     renew
    2.05
     reimburse
    1.65
     repay
    1.56
     inaug
    1.51
     portray
    1.49
     divest
    1.43
     indict
    1.39
     entitle
    1.28
     substant
    1.22
     resign
    1.22
    Activations Density 0.027%

    No Known Activations