© 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. 3-RES-JB
    11. 6348
    Prev
    Next
    INDEX
    Explanations

    references to the Centers for Disease Control and Prevention (CDC)

    oai_token-act-pair · gpt-3.5-turbo

    references to the Centers for Disease Control and Prevention (CDC)

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     istg
    -0.82
    wagen
    -0.74
    ighters
    -0.72
    coni
    -0.69
    othes
    -0.67
     srfAttach
    -0.67
    Redditor
    -0.67
    edom
    -0.66
     hrs
    -0.66
    othal
    -0.65
    POSITIVE LOGITS
    CDC
    1.04
    alse
    0.87
     CDC
    0.86
    rix
    0.74
     Outbreak
    0.72
     Infect
    0.70
     microbi
    0.68
     outbreak
    0.67
    ERG
    0.66
     epidem
    0.66
    Activations Density 0.007%

    No Known Activations