© 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. 11-RES-JB
    11. 19317
    Prev
    Next
    INDEX
    Explanations

    phrases indicating a shift in focus or exploration of different topics

    oai_token-act-pair · gpt-3.5-turbo

    instances of the phrase "a" or similar structures indicating pauses or transitions in discussions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    APD
    -0.75
    acid
    -0.72
    anism
    -0.64
    osity
    -0.63
    arsity
    -0.62
    ocated
    -0.62
    unknown
    -0.60
    avan
    -0.60
    alse
    -0.59
    ted
    -0.59
    POSITIVE LOGITS
     moment
    1.08
     bit
    1.08
     little
    1.07
     refres
    1.07
     few
    1.03
     couple
    0.99
     quick
    0.96
     closer
    0.95
     minute
    0.94
     disclaimer
    0.92
    Activations Density 0.171%

    No Known Activations