© 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. 20474
    Prev
    Next
    INDEX
    Explanations

    words related to elevated structures or landscapes

    oai_token-act-pair · gpt-3.5-turbo

    references to outdoor spaces or terraces

    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
    ittee
    -0.74
     Brotherhood
    -0.74
    gans
    -0.68
    umen
    -0.66
     Fein
    -0.66
     Gadget
    -0.65
     Wass
    -0.65
    Chicago
    -0.64
     Macy
    -0.64
     Key
    -0.64
    POSITIVE LOGITS
     terr
    3.76
     Terr
    1.82
     horr
    1.60
     Tyr
    1.12
     fasc
    1.09
     barr
    1.03
     frust
    1.00
     terrain
    0.98
     patio
    0.96
     rav
    0.93
    Activations Density 0.028%

    No Known Activations