© 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. 5068
    Prev
    Next
    INDEX
    Explanations

    mention of the sun

    oai_token-act-pair · gpt-3.5-turbo

    references to the sun

    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
     Mellon
    -1.04
    elling
    -0.72
    ack
    -0.71
    Ware
    -0.69
    abilia
    -0.68
    ENCY
    -0.66
    Community
    -0.66
    worthy
    -0.65
     Doyle
    -0.65
    iciary
    -0.65
    POSITIVE LOGITS
     sun
    3.77
     sunlight
    2.23
    sun
    2.20
     Sun
    2.17
     sunshine
    2.00
    Sun
    1.92
     sunrise
    1.66
     solar
    1.59
     sunset
    1.52
     sunscreen
    1.51
    Activations Density 0.017%

    No Known Activations