© 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. 23316
    Prev
    Next
    INDEX
    Explanations

    closures, likely related to the end of articles or sections within a document

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "CLOSE"

    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
     Orient
    -0.68
     Sheikh
    -0.65
     Mu
    -0.64
     Adrian
    -0.64
    opl
    -0.63
     Lab
    -0.63
     Mor
    -0.61
     Mag
    -0.61
     Nin
    -0.61
    ochond
    -0.60
    POSITIVE LOGITS
    CLOSE
    4.96
    SHARE
    1.48
    âĸº
    1.22
     WATCHED
    1.21
    PHOTOS
    1.18
    Enlarge
    1.18
    FILE
    1.14
    TIME
    1.12
    WATCH
    1.11
    close
    1.08
    Activations Density 0.014%

    No Known Activations