© 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. 1-RES-JB
    11. 14124
    Prev
    Next
    INDEX
    Explanations

    phrases indicating a strong interest or enthusiasm

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "keen," indicating a focus on enthusiasm or interest

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    代
    -0.82
    adr
    -0.66
    iop
    -0.66
    misc
    -0.65
    ICAN
    -0.65
    anders
    -0.63
    utions
    -0.63
     Incarn
    -0.63
    TPS
    -0.61
    ibi
    -0.60
    POSITIVE LOGITS
     keen
    1.12
    emouth
    0.90
    eful
    0.88
    entimes
    0.87
    ly
    0.85
    ness
    0.83
     Interest
    0.80
    irection
    0.78
     enough
    0.77
     interest
    0.76
    Activations Density 0.005%

    No Known Activations