© 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. 8-RES-JB
    11. 2567
    Prev
    Next
    INDEX
    Explanations

    numerical expressions indicating a quantity of "one and a half."

    oai_token-act-pair · gpt-3.5-turbo

    phrases containing repeated phrases or patterns, particularly those involving the word "and" with related numerical quantities, indicating a focus on conjunctions and numeric expressions

    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
    ancial
    -0.70
    udence
    -0.61
    gen
    -0.60
    insula
    -0.59
    dayName
    -0.59
    rawdownloadcloneembedreportprint
    -0.58
    Gaza
    -0.58
    MRI
    -0.57
    CLA
    -0.57
    omi
    -0.56
    POSITIVE LOGITS
    rogens
    0.96
    rogen
    0.89
    rew
    0.69
     then
    0.68
     counting
    0.68
    romeda
    0.67
     therefore
    0.67
     consequently
    0.67
    seless
    0.66
    tons
    0.65
    Activations Density 0.106%

    No Known Activations