Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 5438
    Prev
    Next
    INDEX
    Explanations

    numerical values representing currency amounts

    oai_token-act-pair · gpt-3.5-turbo

    numerical values or amounts, particularly those related to currency or pricing

    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
     POLITICO
    -0.76
    iven
    -0.72
    orn
    -0.69
    orns
    -0.67
     Robert
    -0.65
     Viol
    -0.61
     Benedict
    -0.61
     misconduct
    -0.60
    nec
    -0.59
     ed
    -0.58
    POSITIVE LOGITS
    5000
    2.97
    4000
    2.92
    3000
    2.85
     3000
    2.70
    6000
    2.60
    8000
    2.58
     5000
    2.57
     4000
    2.50
     6000
    2.47
     8000
    2.43
    Activations Density 0.016%

    No Known Activations