© 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. 6185
    Prev
    Next
    INDEX
    Explanations

    phrases indicating the current situation or time

    oai_token-act-pair · gpt-3.5-turbo

    instances of the phrase "Right now."

    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
    è¦ļéĨĴ
    -0.63
     cellul
    -0.60
    CLUS
    -0.60
    REC
    -0.59
    odium
    -0.56
    raved
    -0.54
     Remastered
    -0.54
    ITED
    -0.54
    anchez
    -0.53
    âĢ¢
    -0.53
    POSITIVE LOGITS
    eous
    1.35
     now
    1.06
    move
    0.85
     fielder
    0.85
     smack
    0.85
     wing
    0.84
    fully
    0.82
    wing
    0.80
     away
    0.80
    now
    0.77
    Activations Density 0.029%

    No Known Activations