© 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. 10-RES-JB
    11. 1417
    Prev
    Next
    INDEX
    Explanations

    instances where there is physical or metaphorical "room for improvement" or expansion

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating the presence of room for improvement or potential

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    book
    -0.79
    jab
    -0.78
    borough
    -0.72
    loo
    -0.72
    til
    -0.70
    edin
    -0.69
    mare
    -0.68
    soever
    -0.68
    fleet
    -0.68
    cles
    -0.67
    POSITIVE LOGITS
     improvement
    1.28
     experimentation
    1.25
     advancement
    1.06
     negotiation
    1.04
     refinement
    1.03
     creativity
    1.02
     exploration
    0.96
     improvis
    0.95
     redemption
    0.94
     contemplation
    0.93
    Activations Density 0.180%

    No Known Activations