© 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. 0-RES-JB
    11. 6710
    Prev
    Next
    INDEX
    Explanations

    negative statements or situations

    oai_token-act-pair · gpt-3.5-turbo

    references to the concept of "nothing."

    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
    ally
    -0.72
    ember
    -0.71
    ural
    -0.68
    urally
    -0.67
    atal
    -0.64
    eal
    -0.62
    sg
    -0.62
    âĢIJ
    -0.61
    ®
    -0.60
    ationally
    -0.57
    POSITIVE LOGITS
     nothing
    3.32
    nothing
    2.52
    Nothing
    2.12
     Nothing
    2.07
     nowhere
    1.82
     none
    1.64
     nobody
    1.52
     anything
    1.49
     everything
    1.44
    everything
    1.34
    Activations Density 0.016%

    No Known Activations