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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 16630
    Prev
    Next
    INDEX
    Explanations

    minerals or mineral-related terms

    oai_token-act-pair · gpt-3.5-turbo

    references to specific minerals and their properties

    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
     Huff
    -0.81
    Buzz
    -0.77
     Squad
    -0.76
     Wilde
    -0.75
     NYPD
    -0.75
    Pad
    -0.70
    oldemort
    -0.70
     Fare
    -0.70
     Suarez
    -0.69
     Finch
    -0.69
    POSITIVE LOGITS
     minerals
    3.65
     mineral
    3.43
     Mineral
    2.62
     Miner
    2.26
     mining
    1.61
     mined
    1.60
     metals
    1.53
     Geological
    1.49
     Mining
    1.46
     miner
    1.45
    Activations Density 0.021%

    No Known Activations