© 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. 5361
    Prev
    Next
    INDEX
    Explanations

    mentions of actions or capabilities of objects being described

    oai_token-act-pair · gpt-3.5-turbo

    instances of the verb "be" in various forms

    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
     Dare
    -0.74
     Matters
    -0.70
     Hail
    -0.70
    Alright
    -0.65
    tery
    -0.65
    orously
    -0.64
     inev
    -0.63
    forcing
    -0.63
    ting
    -0.62
     Promise
    -0.61
    POSITIVE LOGITS
     reused
    1.45
     accessed
    1.43
     interacted
    1.39
     used
    1.39
     swapped
    1.35
     utilized
    1.31
     exchanged
    1.25
     manipulated
    1.19
     searched
    1.19
     traded
    1.18
    Activations Density 0.095%

    No Known Activations