© 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. 18164
    Prev
    Next
    INDEX
    Explanations

    phrases related to companies, contracts, and business partnerships

    oai_token-act-pair · gpt-3.5-turbo

    mentions of consulting firms or associates

    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
    ä¹
    -0.69
    Fu
    -0.64
     pig
    -0.63
     rabbit
    -0.62
     hero
    -0.62
     torn
    -0.60
     Holo
    -0.60
     rider
    -0.59
     pope
    -0.59
    uvian
    -0.59
    POSITIVE LOGITS
     Associates
    4.35
     Consulting
    1.45
     Advis
    1.22
     associates
    1.21
     Architects
    1.16
     Associate
    1.13
     Enterprises
    1.06
     Sons
    1.06
     Publishers
    1.04
     Laboratories
    1.02
    Activations Density 0.017%

    No Known Activations