© 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. 425
    Prev
    Next
    INDEX
    Explanations

    words related to fantasy creatures, specifically elves

    oai_token-act-pair · gpt-3.5-turbo

    references to fantasy creatures, specifically elves and orcs

    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
     rapp
    -0.80
    ction
    -0.75
    apeake
    -0.74
     ACS
    -0.73
    anton
    -0.73
    Mos
    -0.72
    quant
    -0.72
     Pap
    -0.71
    iton
    -0.70
    âģ
    -0.69
    POSITIVE LOGITS
     Elf
    3.63
     elf
    3.43
     elves
    3.04
     Elves
    3.04
    Elf
    2.80
     Elven
    2.46
     Fairy
    1.69
     fairy
    1.66
     Dwarf
    1.65
     Orc
    1.63
    Activations Density 0.030%

    No Known Activations