© 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. 0-RES-JB
    11. 11939
    Prev
    Next
    INDEX
    Explanations

    words related to firearms and ammunition, especially focusing on projectile weapons

    oai_token-act-pair · gpt-3.5-turbo

    different forms of the word "spell" and its various usages

    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
     lin
    -0.65
     bas
    -0.63
     sho
    -0.62
     curves
    -0.61
     edges
    -0.61
     Holmes
    -0.61
    Tx
    -0.60
     Jackets
    -0.60
     slopes
    -0.59
     Committees
    -0.58
    POSITIVE LOGITS
    pelling
    4.10
    pelled
    3.89
    pell
    3.36
    pel
    1.44
    uggest
    1.05
    PLIC
    1.03
     propelled
    1.03
     propel
    1.02
    pring
    0.98
     dispel
    0.95
    Activations Density 0.023%

    No Known Activations