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. 1047
    Prev
    Next
    INDEX
    Explanations

    contractions of "should not"

    oai_token-act-pair · gpt-3.5-turbo

    repeated phrases indicating prohibitions or things that should not occur

    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
     CO
    -0.59
     Tamil
    -0.56
    ickets
    -0.55
    ipes
    -0.53
     air
    -0.51
     See
    -0.50
     operational
    -0.49
     avail
    -0.49
     memory
    -0.49
    gment
    -0.48
    POSITIVE LOGITS
     shouldn
    3.36
     should
    2.00
     ought
    2.00
    should
    1.79
     SHOULD
    1.74
     wouldn
    1.71
     isn
    1.55
     deserves
    1.50
     couldn
    1.49
     aren
    1.48
    Activations Density 0.009%

    No Known Activations