© 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. 8-RES-JB
    11. 179
    Prev
    Next
    INDEX
    Explanations

    phrases containing the word "Eng" or "Engagement."

    oai_token-act-pair · gpt-3.5-turbo

    instances of the term "Eng" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    IFIED
    -0.79
    pmwiki
    -0.79
    TPPStreamerBot
    -0.77
    DragonMagazine
    -0.71
     millenn
    -0.70
     Skydragon
    -0.69
     Leilan
    -0.68
    istically
    -0.68
    ©¶æ
    -0.65
     ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³
    -0.65
    POSITIVE LOGITS
    ravings
    1.15
    raved
    1.15
    lishes
    1.14
    rish
    1.10
    agement
    1.06
    lund
    0.98
    iment
    0.96
    rams
    0.95
    ulf
    0.91
    rave
    0.91
    Activations Density 0.033%

    No Known Activations