© 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. 0-RES-JB
    11. 11621
    Prev
    Next
    INDEX
    Explanations

    the word "Marshall"

    oai_token-act-pair · gpt-3.5-turbo

    the name "Marshall" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     fres
    -0.74
    cheat
    -0.73
     arte
    -0.72
    hene
    -0.69
    Path
    -0.67
     AVG
    -0.66
    hes
    -0.65
     PP
    -0.65
    ng
    -0.64
     dil
    -0.63
    POSITIVE LOGITS
     Marshall
    4.05
    Marsh
    2.63
     Marshal
    1.80
     Marsh
    1.47
     Truman
    1.21
     Barney
    1.18
     McDonnell
    1.17
     marsh
    1.13
     Butler
    1.12
     Howard
    1.10
    Activations Density 0.019%

    No Known Activations