© 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. 1-RES-JB
    11. 1996
    Prev
    Next
    INDEX
    Explanations

    phrases related to the action of handling or managing something

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "handle."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ardy
    -0.71
    ult
    -0.67
    useum
    -0.67
    ulsion
    -0.67
    raction
    -0.67
    rav
    -0.65
    urt
    -0.65
    oration
    -0.64
    arthed
    -0.63
    rave
    -0.63
    POSITIVE LOGITS
    bars
    1.38
     handle
    1.32
     handles
    1.31
     Handle
    1.19
    handle
    1.09
    Handle
    0.90
    BOOK
    0.82
    handled
    0.80
    bar
    0.80
    beit
    0.78
    Activations Density 0.007%

    No Known Activations