© 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. 4753
    Prev
    Next
    INDEX
    Explanations

    references to characters, episodes, and events from the television show "Doctor Who."

    oai_token-act-pair · gpt-3.5-turbo

    elements related to the "Doctor Who" franchise, particularly mentions of the Tenth Doctor and associated characters or storylines

    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
     Asheville
    -0.98
     Tulsa
    -0.98
    NBA
    -0.96
     Omaha
    -0.91
     Akron
    -0.91
    asketball
    -0.91
     basketball
    -0.90
     volleyball
    -0.90
    orgetown
    -0.84
     LeBron
    -0.82
    POSITIVE LOGITS
     Doctor
    1.91
     Moff
    1.81
    Doctor
    1.59
     Doctors
    1.47
    ARDIS
    1.42
     Clara
    1.39
     Sherlock
    1.32
     regeneration
    1.25
     Dale
    1.21
    aldi
    1.16
    Activations Density 0.151%

    No Known Activations