© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 11-RES-JB
    11. 20807
    Prev
    Next
    INDEX
    Explanations

    negative adjectives starting with 'un-' followed by descriptive words

    oai_token-act-pair · gpt-3.5-turbo

    negative prefixes, particularly "un," to identify words associated with negativity or absence

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     rows
    -0.78
     flats
    -0.75
     periphery
    -0.72
     racks
    -0.71
     bruises
    -0.70
     indoors
    -0.70
     chained
    -0.70
     separately
    -0.70
     stalls
    -0.70
     deletion
    -0.69
    POSITIVE LOGITS
    help
    1.48
    productive
    1.42
    balanced
    1.38
    important
    1.35
    interesting
    1.32
    professional
    1.30
    readable
    1.27
    original
    1.27
    inspired
    1.26
    ruly
    1.26
    Activations Density 0.029%

    No Known Activations