© 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. 9338
    Prev
    Next
    INDEX
    Explanations

    instances where a concept or quality is described as being equal or equivalent to another concept or quality

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "equally" with varying contexts

    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
    uese
    -0.67
    ourke
    -0.67
    ocene
    -0.63
    hra
    -0.62
    pez
    -0.62
    Äĩ
    -0.62
    aques
    -0.61
     Pros
    -0.61
    ULE
    -0.61
    amily
    -0.61
    POSITIVE LOGITS
     situated
    1.22
     sized
    1.07
     spaced
    0.99
     important
    0.94
     applicable
    0.93
     minded
    0.92
     adept
    0.90
     worthless
    0.86
     weighted
    0.85
     valuable
    0.85
    Activations Density 0.047%

    No Known Activations