© 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. 9-RES-JB
    11. 13291
    Prev
    Next
    INDEX
    Explanations

    words related to scientific studies and their findings

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to scientific studies and their findings

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 9-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    9
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     UNCLASSIFIED
    -0.82
     thereafter
    -0.82
    emort
    -0.78
     continuity
    -0.77
    ultane
    -0.77
    usha
    -0.68
     substantive
    -0.66
     deleg
    -0.66
    ledged
    -0.65
    istical
    -0.65
    POSITIVE LOGITS
    Researchers
    1.37
     researchers
    1.25
     Researchers
    1.21
    Redditor
    1.15
    Scientists
    1.13
     scientists
    1.05
    utterstock
    1.02
     Scientists
    1.00
    Experts
    0.98
     researcher
    0.96
    Activations Density 0.753%

    No Known Activations