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. 11-RES-JB
    6. 10818
    Prev
    Next
    INDEX
    Explanations

    scientific findings or research results

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to research findings and reported behaviors

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    onica
    -0.75
    quote
    -0.69
    onto
    -0.69
    steel
    -0.68
    oland
    -0.68
    undo
    -0.68
    united
    -0.68
    prep
    -0.67
    raft
    -0.67
    icus
    -0.66
    POSITIVE LOGITS
     slight
    1.30
     reductions
    1.26
     negligible
    1.26
     significant
    1.26
     declines
    1.25
     decreases
    1.21
     fewer
    1.21
     minimal
    1.19
     substantial
    1.18
     decreased
    1.18
    Activations Density 0.296%

    No Known Activations