Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 3-RES-JB
    6. 10095
    Prev
    Next
    INDEX
    Explanations

    company names and organizations in the business and technology sector

    oai_token-act-pair · gpt-3.5-turbo

    corporate and brand names within business contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     poems
    -0.82
     bapt
    -0.78
     Trayvon
    -0.75
     recess
    -0.73
     poem
    -0.72
     Caleb
    -0.71
     canon
    -0.71
     Miranda
    -0.70
     Cody
    -0.70
     verse
    -0.70
    POSITIVE LOGITS
     Companies
    1.21
     companies
    1.18
    Companies
    1.17
     Holdings
    1.16
    Company
    1.15
     subsidiaries
    1.15
     firms
    1.13
     pricing
    1.09
     suppliers
    1.08
    Sales
    1.05
    Activations Density 0.574%

    No Known Activations