© 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. 483
    Prev
    Next
    INDEX
    Explanations

    mentions of political offices or events like "Senate" or "campaign trail."

    oai_token-act-pair · gpt-3.5-turbo

    instances of specific references to "the" in a political context

    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
    sterdam
    -0.73
    river
    -0.70
    esters
    -0.69
     reckoned
    -0.67
    nav
    -0.66
    DER
    -0.66
    bour
    -0.65
    azard
    -0.64
    pai
    -0.64
     guessed
    -0.64
    POSITIVE LOGITS
     podium
    1.42
     Oval
    1.19
     stump
    1.08
     Senate
    1.06
     microphone
    1.05
     microphones
    1.00
     pul
    0.98
     sidelines
    0.98
     floor
    0.96
     aisle
    0.95
    Activations Density 0.170%

    No Known Activations