© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 16983
    Prev
    Next
    INDEX
    Explanations

    proper nouns related to specific movies or books, especially "The Force Awakens" and "Dunkirk", as well as some other titles

    oai_token-act-pair · gpt-3.5-turbo

    titles and names of popular movies or franchises

    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
     Colt
    -0.66
     physicians
    -0.66
     rig
    -0.65
     Mitchell
    -0.65
     propri
    -0.65
     hem
    -0.64
     professional
    -0.64
     prof
    -0.64
     scrub
    -0.63
     arm
    -0.62
    POSITIVE LOGITS
     Awakens
    4.44
     Leia
    1.34
     Furious
    1.27
    assic
    1.26
     Rowling
    1.25
     Ultron
    1.20
     Thrones
    1.19
     Unleashed
    1.18
     Skywalker
    1.17
     Hunger
    1.16
    Activations Density 0.029%

    No Known Activations