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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 8-RES-JB
    11. 22681
    Prev
    Next
    INDEX
    Explanations

    terms related to various societal and technical domains, such as traffic management, land faults, policing, surveillance, health outcomes, education, security, music, food, and parenting

    oai_token-act-pair · gpt-3.5-turbo

    key terms related to various societal and environmental issues

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    urat
    -0.64
    alion
    -0.62
    ĪĴ
    -0.60
    ortium
    -0.60
    atever
    -0.59
    wered
    -0.57
    enegger
    -0.56
    ueless
    -0.55
    ALSE
    -0.52
    ģ«
    -0.52
    POSITIVE LOGITS
     âĢº
    0.68
     environments
    0.67
     outcomes
    0.63
     automation
    0.61
     globally
    0.60
     disorders
    0.59
     worldwide
    0.58
     cannabis
    0.58
     establishments
    0.58
     surfaces
    0.57
    Activations Density 0.677%

    No Known Activations