© 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. 16864
    Prev
    Next
    INDEX
    Explanations

    references to specific numerical values, particularly the number 10

    oai_token-act-pair · gpt-3.5-turbo

    references to the number ten

    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
    •Embedding Space Digit Features

    No Comments

    Negative Logits
    ault
    -0.69
    sers
    -0.64
     Canaver
    -0.62
    aults
    -0.59
     defic
    -0.57
     conflic
    -0.57
    essors
    -0.57
    ãĤ«
    -0.57
     orthodoxy
    -0.56
     predec
    -0.56
    POSITIVE LOGITS
     10
    2.78
     11
    2.08
     12
    2.01
     15
    2.00
     20
    1.92
     9
    1.85
     8
    1.82
     14
    1.82
     25
    1.80
     13
    1.80
    Activations Density 0.037%

    No Known Activations