© 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. 11-RES-JB
    11. 7510
    Prev
    Next
    INDEX
    Explanations

    phrases or words that start with 'wh'

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "wh," indicating a focus on questions or questioning phrases

    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
     mosaic
    -0.65
     silenced
    -0.63
     enclosed
    -0.63
     condolences
    -0.62
     marking
    -0.61
     dividing
    -0.61
     expressing
    -0.61
     starters
    -0.60
     issuer
    -0.59
     condu
    -0.59
    POSITIVE LOGITS
    omever
    1.69
    irling
    1.46
    ipl
    1.30
    izz
    1.29
    arf
    1.28
    irl
    1.27
    ittle
    1.27
    olen
    1.26
    acky
    1.26
    itt
    1.23
    Activations Density 0.008%

    No Known Activations