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. 3-RES-JB
    6. 10223
    Prev
    Next
    INDEX
    Explanations

    nouns related to immigration

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "emigrant" or related terms

    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
    aird
    -0.74
    DragonMagazine
    -0.74
     Lovecraft
    -0.70
     Elves
    -0.70
    lihood
    -0.69
     Kavanaugh
    -0.69
     Idle
    -0.68
     Kubrick
    -0.68
    WAYS
    -0.67
    UGH
    -0.67
    POSITIVE LOGITS
    igrants
    1.28
    igrant
    1.21
    igration
    1.19
    igrated
    1.18
    blems
    1.15
    bers
    1.12
    perors
    1.12
    otions
    1.05
    igr
    1.04
    ulsion
    1.04
    Activations Density 0.013%

    No Known Activations