© 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. 20014
    Prev
    Next
    INDEX
    Explanations

    phrases indicating logical conclusions or implications

    oai_token-act-pair · gpt-3.5-turbo

    conjunctions or phrases that indicate causality or conclusions

    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
     dozen
    -0.68
     Edge
    -0.52
     meg
    -0.51
     downtown
    -0.51
    bury
    -0.51
    amer
    -0.50
    uggets
    -0.50
    tone
    -0.49
    attled
    -0.49
     Pavilion
    -0.49
    POSITIVE LOGITS
     therefore
    3.89
     consequently
    2.77
     thus
    2.51
     furthermore
    2.43
     moreover
    2.40
     hence
    2.35
     accordingly
    2.12
    Therefore
    1.93
     Therefore
    1.91
     thereby
    1.82
    Activations Density 0.012%

    No Known Activations