© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 2-RES-JB
    11. 347
    Prev
    Next
    INDEX
    Explanations

    the word "regret" and its different conjugations

    oai_token-act-pair · gpt-4-turbo

    expressions of remorse or apology

    oai_token-act-pair · gpt-3.5-turbo

    expressions of remorse or sadness

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    iop
    -0.71
    ammy
    -0.70
    IPS
    -0.67
    icle
    -0.67
    ymph
    -0.65
    chool
    -0.65
    akedown
    -0.65
    ipt
    -0.64
    anto
    -0.64
     dotted
    -0.63
    POSITIVE LOGITS
     regrets
    1.08
     regret
    1.04
    fully
    0.99
     regretted
    0.94
    imaru
    0.88
    fulness
    0.87
    ful
    0.86
     remorse
    0.84
    faced
    0.82
    rence
    0.81
    Activations Density 0.004%

    No Known Activations