© 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. 8-RES-JB
    11. 1331
    Prev
    Next
    INDEX
    Explanations

    phrases relating to comparisons or contrasts between two entities

    oai_token-act-pair · gpt-3.5-turbo

    references to comparisons between two entities or groups

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    vich
    -0.84
    nell
    -0.81
    ucked
    -0.78
    terday
    -0.75
    renheit
    -0.75
    daq
    -0.74
    owler
    -0.74
    levard
    -0.73
    pletion
    -0.72
    lehem
    -0.71
    POSITIVE LOGITS
     extremes
    1.30
     sexes
    1.27
     halves
    1.21
     sides
    1.14
     genders
    1.06
     parties
    1.02
     worlds
    0.93
     realms
    0.93
     poles
    0.91
     universes
    0.88
    Activations Density 0.101%

    No Known Activations