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. 4286
    Prev
    Next
    INDEX
    Explanations

    scientific results and research findings

    oai_token-act-pair · gpt-3.5-turbo

    results and findings from scientific studies or research

    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
     Vanity
    -0.79
     Kits
    -0.73
     Citizenship
    -0.70
     Exile
    -0.70
     bankrupt
    -0.69
     pilgr
    -0.66
    ãĤ¢ãĥ«
    -0.65
     Citizen
    -0.64
    ãĤ½
    -0.62
     Collector
    -0.62
    POSITIVE LOGITS
    CONCLUS
    1.32
    suggest
    1.30
    Interestingly
    1.15
     suggests
    1.14
     suggesting
    1.13
     hypothes
    1.08
     hypothesized
    1.07
     suggestive
    1.03
     suggest
    1.02
     conclusion
    1.02
    Activations Density 0.309%

    No Known Activations