© 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. 4-RES-JB
    11. 2624
    Prev
    Next
    INDEX
    Explanations

    statements where comments are declined

    oai_token-act-pair · gpt-3.5-turbo

    instances of the phrase "declined to comment."

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

    No Comments

    Negative Logits
     Sinai
    -0.75
    ppo
    -0.73
     Ethiopian
    -0.68
    nown
    -0.66
     pilgr
    -0.65
    hew
    -0.65
    Sensor
    -0.64
    tail
    -0.63
    married
    -0.63
    locked
    -0.63
    POSITIVE LOGITS
    ariat
    1.01
    ysis
    0.89
    aries
    0.82
    ispers
    0.78
    aires
    0.78
    ively
    0.78
    ivity
    0.77
     ado
    0.75
    ault
    0.75
    ode
    0.74
    Activations Density 0.025%

    No Known Activations