© 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. 8019
    Prev
    Next
    INDEX
    Explanations

    instances of words related to competing or recipes

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to competition and recipes

    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
     flags
    -0.73
     lightning
    -0.68
     CLSID
    -0.65
     pas
    -0.65
     pacing
    -0.64
     unit
    -0.63
     delegation
    -0.63
     ding
    -0.63
     progressives
    -0.62
     Wesley
    -0.62
    POSITIVE LOGITS
    Recipe
    2.74
     competed
    1.82
     Recipe
    1.65
    Shop
    1.07
    Cub
    1.07
     Marin
    0.99
    Nut
    0.97
     Bake
    0.95
    Serv
    0.92
    effective
    0.90
    Activations Density 0.043%

    No Known Activations