© 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. 11376
    Prev
    Next
    INDEX
    Explanations

    contractions, specifically those in which 's is used as a contraction for "is" or "has"

    oai_token-act-pair · gpt-3.5-turbo

    phrases with the structure "It’s [adjective/phrase]" indicating subjective opinions or observations

    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
    rongh
    -0.72
    76561
    -0.70
    luaj
    -0.70
    phe
    -0.64
     Bearing
    -0.62
    continue
    -0.62
    rite
    -0.61
    rang
    -0.61
     ({
    -0.61
    ONG
    -0.60
    POSITIVE LOGITS
     ironic
    1.06
     tempting
    1.04
     unclear
    1.02
     easy
    1.00
     why
    0.98
     impossible
    0.97
     easier
    0.97
     understandable
    0.93
     imperative
    0.93
     doubtful
    0.89
    Activations Density 0.170%

    No Known Activations