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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 11-RES-JB
    11. 5893
    Prev
    Next
    INDEX
    Explanations

    references to blog posts and news articles

    oai_token-act-pair · gpt-3.5-turbo

    keywords related to sharing and promoting content

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    sic
    -0.73
    <|endoftext|>
    -0.64
    anwhile
    -0.61
    terday
    -0.61
     UNCLASSIFIED
    -0.59
    Advertisements
    -0.57
    cember
    -0.53
    forth
    -0.53
     remorse
    -0.52
    pled
    -0.52
    POSITIVE LOGITS
    Phys
    0.60
    Streamer
    0.60
     âĢº
    0.59
     Manga
    0.54
    catentry
    0.52
    Show
    0.52
    duration
    0.52
    wiki
    0.52
     Event
    0.51
     Blend
    0.50
    Activations Density 0.205%

    No Known Activations