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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 10087
    Prev
    Next
    INDEX
    Explanations

    words related to being confused

    oai_token-act-pair · gpt-3.5-turbo

    expressions of confusion

    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
    bay
    -0.78
    ano
    -0.75
    Ro
    -0.71
    fare
    -0.67
    anol
    -0.66
    saf
    -0.65
     Materials
    -0.65
    aro
    -0.65
    AGE
    -0.65
    ofi
    -0.64
    POSITIVE LOGITS
     confused
    3.53
     puzzled
    2.10
     baffled
    2.01
     confuse
    1.89
     confusion
    1.89
     mistaken
    1.88
     startled
    1.67
     perplex
    1.67
     conflicted
    1.63
     frightened
    1.62
    Activations Density 0.021%

    No Known Activations