© 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. 7961
    Prev
    Next
    INDEX
    Explanations

    words related to espionage and intelligence operations

    oai_token-act-pair · gpt-3.5-turbo

    repeated mentions of the term "esp."

    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
     Kai
    -0.69
     closed
    -0.68
     Territory
    -0.65
     duck
    -0.64
     Duel
    -0.64
     Kan
    -0.63
     diagram
    -0.63
     clarity
    -0.62
     typed
    -0.60
     Ot
    -0.60
    POSITIVE LOGITS
    esp
    4.98
    esm
    2.51
     esp
    1.54
    psc
    1.35
     ESP
    1.32
    yrim
    1.15
    ersive
    1.13
     Skyrim
    1.13
    ovie
    1.04
    esc
    1.03
    Activations Density 0.021%

    No Known Activations