© 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. 11-RES-JB
    11. 18770
    Prev
    Next
    INDEX
    Explanations

    sentences related to medical research or studies, particularly focused on smoking cessation programs and medical advancements

    oai_token-act-pair · gpt-3.5-turbo

    punctuation and sentence structure, primarily focusing on full stops and commas

    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
     grievance
    -0.81
     plead
    -0.80
     welf
    -0.78
     deterrent
    -0.77
     certific
    -0.76
    abouts
    -0.76
     condos
    -0.75
     buyer
    -0.74
     barg
    -0.74
     occup
    -0.74
    POSITIVE LOGITS
     Researchers
    1.54
    Researchers
    1.41
     Scientists
    1.35
     Previous
    1.25
    Scientists
    1.22
    Study
    1.20
     Their
    1.17
    Professor
    1.17
    Previous
    1.15
     Authors
    1.14
    Activations Density 0.261%

    No Known Activations