© 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. 554
    Prev
    Next
    INDEX
    Explanations

    security-related terms such as "cam," "admin," and "password."

    oai_token-act-pair · gpt-3.5-turbo

    references to surveillance technology and administrative roles

    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
    ģĸ
    -0.76
     generational
    -0.76
     resil
    -0.71
     antioxidants
    -0.71
     Twain
    -0.70
     Gust
    -0.69
    Enlarge
    -0.68
     Mavericks
    -0.68
     Revolutionary
    -0.67
    eteenth
    -0.65
    POSITIVE LOGITS
    cam
    2.81
     admin
    2.11
    admin
    1.98
     Admin
    1.90
    Admin
    1.84
     Ops
    1.71
     medic
    1.15
     ops
    1.14
     admins
    1.13
    Ops
    1.06
    Activations Density 0.053%

    No Known Activations