© 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. 18997
    Prev
    Next
    INDEX
    Explanations

    phrases with contractions indicating inability or negation

    oai_token-act-pair · gpt-3.5-turbo

    the word "couldn’t" in various contexts

    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
     normative
    -0.64
     appropriate
    -0.57
     free
    -0.57
     current
    -0.55
     accordance
    -0.55
     similar
    -0.55
     reciprocal
    -0.55
     Messages
    -0.55
     reporting
    -0.52
    gress
    -0.52
    POSITIVE LOGITS
     couldn
    3.10
     wouldn
    2.08
     didn
    1.98
     hadn
    1.92
     wasn
    1.82
     weren
    1.79
     shouldn
    1.72
    didn
    1.72
     cannot
    1.64
    could
    1.63
    Activations Density 0.011%

    No Known Activations