© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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. 16820
    Prev
    Next
    INDEX
    Explanations

    statements related to policies, laws, and regulations

    oai_token-act-pair · gpt-3.5-turbo

    instances of the phrase "In" indicating the beginning of a new thought or point in the text

    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
    Redditor
    -0.75
     Gorge
    -0.67
    hedon
    -0.67
     scares
    -0.66
    ruciating
    -0.64
    Enlarge
    -0.64
     humiliated
    -0.64
     crap
    -0.64
     spoil
    -0.63
     roared
    -0.63
    POSITIVE LOGITS
     addition
    1.28
     conclusion
    1.27
    clusion
    1.21
     summary
    1.16
    strument
    1.13
    sofar
    1.11
    ventory
    1.10
     particular
    1.08
     contrast
    1.03
    cluded
    1.03
    Activations Density 0.142%

    No Known Activations