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

    Neuronpedia

    Jacobian LensNEW
    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. 1928
    Prev
    Next
    INDEX
    Explanations

    words related to approval or positive reception

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "welcomed."

    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
    sever
    -0.66
     belt
    -0.62
     corrupt
    -0.61
     obsc
    -0.59
    plan
    -0.58
    fail
    -0.56
     Unreal
    -0.56
    letes
    -0.55
     screw
    -0.55
     vest
    -0.55
    POSITIVE LOGITS
     welcomed
    3.71
     welcome
    2.15
     greeted
    1.97
     welcomes
    1.96
     applauded
    1.85
     appreciated
    1.57
     hailed
    1.57
     cheered
    1.53
     embraced
    1.52
     congratulated
    1.45
    Activations Density 0.014%

    No Known Activations