© 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. 3-RES-JB
    11. 3830
    Prev
    Next
    INDEX
    Explanations

    terms related to increases or enhancements

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "increased" and related terms that denote growth or improvement

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Tycoon
    -0.82
    Bio
    -0.81
    verse
    -0.77
    rics
    -0.72
    breaks
    -0.71
    aries
    -0.70
    mates
    -0.70
     ¯
    -0.70
     Impossible
    -0.70
     Outs
    -0.69
    POSITIVE LOGITS
     tremend
    1.09
     srf
    1.03
     satell
    1.01
     carbohyd
    0.99
     earthqu
    0.98
     confir
    0.96
     exha
    0.95
    undai
    0.94
     eleph
    0.94
    ¥ŀ
    0.93
    Activations Density 0.047%

    No Known Activations