© 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. 11-RES-JB
    11. 2247
    Prev
    Next
    INDEX
    Explanations

    words related to negative qualities or conditions such as hopelessness, helplessness, deservingness, and baldness

    oai_token-act-pair · gpt-3.5-turbo

    terms related to various states of being or conditions, often with a focus on negative aspects

    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
    Downloadha
    -0.75
    arty
    -0.72
     irrig
    -0.67
    76561
    -0.67
     DRAG
    -0.63
     sidx
    -0.63
     Mehran
    -0.62
     Strange
    -0.61
     Greenwald
    -0.61
     Cheap
    -0.59
    POSITIVE LOGITS
    ness
    3.19
    nesses
    2.55
    NESS
    2.30
    liness
    1.66
    ity
    1.62
    itude
    1.39
    ening
    1.37
    eness
    1.31
    nes
    1.22
    cies
    1.21
    Activations Density 0.296%

    No Known Activations