© 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. 3-RES-JB
    11. 8140
    Prev
    Next
    INDEX
    Explanations

    words related to emotional and physical comfort

    oai_token-act-pair · gpt-3.5-turbo

    references to "comfort" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    etheus
    -0.77
    sonian
    -0.76
    yss
    -0.72
    natureconservancy
    -0.70
    uscript
    -0.69
    atta
    -0.67
    Torrent
    -0.67
    rama
    -0.66
    degree
    -0.65
    unes
    -0.65
    POSITIVE LOGITS
     comfort
    1.19
     Comfort
    1.10
    comfort
    1.09
    ably
    0.83
    raints
    0.83
     margins
    0.78
    iously
    0.76
     reassure
    0.76
    ãĥ¼ãĥĨ
    0.74
     Seat
    0.74
    Activations Density 0.015%

    No Known Activations