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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 0-RES-JB
    11. 20265
    Prev
    Next
    INDEX
    Explanations

    questions starting with "How" followed by a topic

    oai_token-act-pair · gpt-3.5-turbo

    questions starting with "How."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     recess
    -0.69
     stands
    -0.68
     committed
    -0.66
     designated
    -0.65
     stand
    -0.64
     unspecified
    -0.62
     reneg
    -0.62
     otherwise
    -0.62
     priority
    -0.62
     capital
    -0.62
    POSITIVE LOGITS
    How
    3.01
    Why
    2.13
    how
    2.03
    HOW
    2.00
     How
    1.97
    What
    1.92
    Who
    1.75
     HOW
    1.72
    Where
    1.66
    Does
    1.65
    Activations Density 0.012%

    No Known Activations