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. 0-RES-JB
    6. 3707
    Prev
    Next
    INDEX
    Explanations

    words related to geographical locations or maps

    oai_token-act-pair · gpt-3.5-turbo

    references to maps or mapping concepts

    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
     teeth
    -0.87
     lun
    -0.73
     disbelief
    -0.70
     Russ
    -0.69
     shareholder
    -0.68
     Walters
    -0.67
     rept
    -0.66
     Scientology
    -0.63
     bree
    -0.63
     perjury
    -0.63
    POSITIVE LOGITS
    map
    4.29
    maps
    2.99
    Map
    2.46
    MAP
    2.09
     map
    2.05
     Map
    1.98
    Maps
    1.93
     maps
    1.66
     mapped
    1.62
     mapping
    1.58
    Activations Density 0.012%

    No Known Activations