© 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. 20595
    Prev
    Next
    INDEX
    Explanations

    terms related to the programming tool "Gradle"

    oai_token-act-pair · gpt-3.5-turbo

    references to graduate-level education or grad school concepts

    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
    ħĭ
    -0.76
     Shack
    -0.75
     Cust
    -0.67
    66666666
    -0.66
    BOOK
    -0.64
     Seeking
    -0.64
     Hole
    -0.64
     constitu
    -0.64
     Role
    -0.61
     Queensland
    -0.59
    POSITIVE LOGITS
    uates
    1.47
    uations
    1.29
    uated
    1.27
    uation
    1.26
    uating
    1.18
    ients
    1.13
    uate
    1.12
    iate
    1.02
    ient
    1.00
     grad
    0.88
    Activations Density 0.009%

    No Known Activations