© 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. 1-RES-JB
    11. 8658
    Prev
    Next
    INDEX
    Explanations

    phrases expressing apologies or regrets

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "Sorry" expressing apology or regret

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     helicop
    -0.76
     holistic
    -0.71
    rower
    -0.70
    vet
    -0.69
    ellation
    -0.68
    estial
    -0.68
    anced
    -0.68
    minist
    -0.67
     anchored
    -0.67
     guided
    -0.67
    POSITIVE LOGITS
    Sorry
    1.24
     Sorry
    1.21
     sorry
    1.18
    sorry
    1.16
     Oops
    0.81
     Slug
    0.81
     Guys
    0.80
    éĹĺ
    0.78
     Credits
    0.77
    imaru
    0.77
    Activations Density 0.016%

    No Known Activations