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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 16795
    Prev
    Next
    INDEX
    Explanations

    the word "want" in sentences expressing desires or preferences

    oai_token-act-pair · gpt-3.5-turbo

    expressions of desire or intention

    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
    milo
    -0.76
    forms
    -0.66
    clus
    -0.59
     Bearing
    -0.59
    yne
    -0.58
    former
    -0.55
    buster
    -0.55
     cush
    -0.55
    heid
    -0.55
     borne
    -0.54
    POSITIVE LOGITS
     want
    3.03
    want
    2.44
     wanna
    2.27
     WANT
    2.13
     wants
    2.09
     Want
    2.04
     wanted
    1.88
     wish
    1.80
     wanting
    1.66
     need
    1.58
    Activations Density 0.050%

    No Known Activations