© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 24144
    Prev
    Next
    INDEX
    Explanations

    terms related to espionage or suspicious activities

    oai_token-act-pair · gpt-3.5-turbo

    specific names and terms related to entities, actions, and conditions 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
    ;;
    -0.56
     crappy
    -0.54
    GGGGGGGG
    -0.54
    Ĭ±
    -0.53
     bothering
    -0.52
     Colt
    -0.52
    Huh
    -0.51
    ĪĴ
    -0.51
     veter
    -0.51
     Charleston
    -0.49
    POSITIVE LOGITS
     lasted
    0.85
     lasts
    0.84
     underwent
    0.84
     extends
    0.84
     relies
    0.83
     also
    0.83
     survives
    0.79
     retains
    0.79
     comprises
    0.77
     nevertheless
    0.77
    Activations Density 0.820%

    No Known Activations