© 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. 1-RES-JB
    11. 20642
    Prev
    Next
    INDEX
    Explanations

    references to various types of worms

    oai_token-act-pair · gpt-3.5-turbo

    references to worms and related terminology

    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
    ctic
    -0.79
    ioch
    -0.74
    Court
    -0.69
    ++++++++++++++++
    -0.68
    -----------
    -0.67
    rylic
    -0.64
    owered
    -0.63
    cca
    -0.63
    ulton
    -0.63
    trl
    -0.63
    POSITIVE LOGITS
    worm
    1.40
    worms
    1.37
    hole
    1.33
     worms
    1.26
     worm
    1.20
    holes
    1.10
     Worm
    1.05
    fish
    1.01
    tail
    0.90
     reefs
    0.89
    Activations Density 0.004%

    No Known Activations