© 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. 2-RES-JB
    11. 8386
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "worlds"

    oai_token-act-pair · gpt-3.5-turbo

    references to different fictional realms or settings

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ged
    -0.81
    thumbnails
    -0.77
    ging
    -0.76
    sie
    -0.75
    draw
    -0.69
    onomy
    -0.67
    cer
    -0.65
    uments
    -0.64
    200000
    -0.63
    rav
    -0.63
    POSITIVE LOGITS
     worlds
    1.03
    chool
    0.91
    hops
    0.86
     collide
    0.81
    peak
    0.79
    ervative
    0.78
     Worlds
    0.77
     Ceres
    0.75
     Reborn
    0.75
    afety
    0.74
    Activations Density 0.015%

    No Known Activations