© 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. 0-RES-JB
    11. 18212
    Prev
    Next
    INDEX
    Explanations

    phrases related to coercion or being compelled to do something against one's will

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "forced."

    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
    án
    -0.90
     Review
    -0.70
    ight
    -0.68
    steen
    -0.67
    tern
    -0.66
     Beard
    -0.66
     Principles
    -0.65
    mary
    -0.65
    Mods
    -0.65
    oho
    -0.63
    POSITIVE LOGITS
     forced
    3.38
     Forced
    2.36
    forced
    2.18
     compelled
    2.13
     coerced
    1.97
     forcing
    1.82
     pressured
    1.78
     obliged
    1.48
     forcibly
    1.47
     prevented
    1.37
    Activations Density 0.019%

    No Known Activations