© 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. 512
    Prev
    Next
    INDEX
    Explanations

    references to a person named "Johansen."

    oai_token-act-pair · gpt-3.5-turbo

    the name "Joh" and its variations in the text

    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
     array
    -0.69
     aux
    -0.67
     Doomsday
    -0.67
     theater
    -0.65
    flies
    -0.65
     piled
    -0.64
     holdings
    -0.64
    cap
    -0.63
    arbon
    -0.62
    Army
    -0.61
    POSITIVE LOGITS
     Joh
    3.87
     Johann
    1.52
     Jak
    1.35
     Bj
    1.31
     Jol
    1.24
     Boh
    1.20
     Johannes
    1.18
     Ferdinand
    1.18
     Jah
    1.14
     Rasm
    1.10
    Activations Density 0.025%

    No Known Activations