© 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. 0-RES-JB
    11. 10146
    Prev
    Next
    INDEX
    Explanations

    words related to specific intentions or objectives

    oai_token-act-pair · gpt-3.5-turbo

    references to various kinds of purposes or intended uses

    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
    Domin
    -0.64
    Äį
    -0.63
    line
    -0.62
    obi
    -0.62
    anian
    -0.62
    obin
    -0.61
    former
    -0.60
    rentice
    -0.60
    osa
    -0.60
    panic
    -0.59
    POSITIVE LOGITS
     purposes
    3.88
     purpose
    2.10
     sake
    1.82
     reasons
    1.72
    purpose
    1.41
     Purpose
    1.28
     objectives
    1.19
     motives
    1.17
     meanings
    1.16
    ummies
    1.11
    Activations Density 0.014%

    No Known Activations