© 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. 7-RES-JB
    11. 19467
    Prev
    Next
    INDEX
    Explanations

    ingredients and items commonly associated with recipes and home cooking

    oai_token-act-pair · gpt-3.5-turbo

    terms related to medical conditions, particularly those involving physical health and dietary components

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 7-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    7
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    lyak
    -0.63
    atively
    -0.58
    Redditor
    -0.58
    afer
    -0.57
     answ
    -0.56
    dule
    -0.55
     proble
    -0.54
    !'"
    -0.53
     confir
    -0.53
    icably
    -0.52
    POSITIVE LOGITS
     etc
    1.14
    etc
    1.08
     Miscellaneous
    0.73
     ect
    0.71
     Sleeping
    0.64
     Misc
    0.62
     âĵĺ
    0.60
    whatever
    0.55
    ĪĴ
    0.54
     assorted
    0.52
    Activations Density 0.247%

    No Known Activations