Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 20963
    Prev
    Next
    INDEX
    Explanations

    phrases related to detachment or removal of something

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to the term "off," particularly in various contexts such as instructions, actions, or occurrences

    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
    ocally
    -0.65
     Icar
    -0.62
    Ãį
    -0.61
     pores
    -0.60
     pronouns
    -0.58
     swollen
    -0.57
     awaited
    -0.57
     patience
    -0.56
     commonly
    -0.55
     precious
    -0.55
    POSITIVE LOGITS
    off
    4.14
    offs
    3.18
    OFF
    2.50
    Off
    2.19
     Off
    1.73
    offer
    1.70
     OFF
    1.46
     off
    1.43
    away
    1.25
    out
    1.21
    Activations Density 0.008%

    No Known Activations