© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 12175
    Prev
    Next
    INDEX
    Explanations

    phrases related to older objects, people, or concepts

    oai_token-act-pair · gpt-3.5-turbo

    the word "old" and its variations with varying emphasis

    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
    Resp
    -0.68
     additionally
    -0.65
     respectively
    -0.63
    Click
    -0.63
     Particip
    -0.61
     furthermore
    -0.61
    WN
    -0.61
    scape
    -0.61
    respond
    -0.60
    âĢİ
    -0.60
    POSITIVE LOGITS
     old
    3.53
    Old
    2.25
     olds
    2.10
     older
    2.07
    old
    2.04
     Old
    1.96
     outdated
    1.61
     oldest
    1.61
     ol
    1.54
     ancient
    1.52
    Activations Density 0.024%

    No Known Activations