© 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. 4-RES-JB
    11. 16036
    Prev
    Next
    INDEX
    Explanations

    onomatopoeic words like "barking," words related to physical expressions like "dancing," and expressions of joy like "smiling" and "grinning."

    oai_token-act-pair · gpt-3.5-turbo

    actions associated with positive emotions and energetic behaviors

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

    No Comments

    Negative Logits
    xy
    -0.88
    xus
    -0.82
    yon
    -0.82
    pora
    -0.80
    zanne
    -0.77
    pps
    -0.76
    cms
    -0.75
    mel
    -0.74
    si
    -0.74
    olog
    -0.74
    POSITIVE LOGITS
     bananas
    0.87
     behavi
    0.80
     Lizard
    0.78
     exha
    0.78
    GGGGGGGG
    0.77
     Squid
    0.76
     ducks
    0.76
     gib
    0.75
    NESS
    0.75
     eyeb
    0.72
    Activations Density 0.150%

    No Known Activations