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. 5924
    Prev
    Next
    INDEX
    Explanations

    words related to the word "hammock."

    oai_token-act-pair · gpt-3.5-turbo

    references to specific names and terms related to objects or locations, particularly "Hamm" and "doors."

    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
     Frieza
    -0.85
    clair
    -0.85
    ipop
    -0.78
    emia
    -0.77
     Vet
    -0.77
     Leone
    -0.77
    Card
    -0.75
    ename
    -0.75
    Corp
    -0.73
    eye
    -0.73
    POSITIVE LOGITS
     hamm
    2.84
     Hamm
    2.18
     padd
    1.91
     raft
    1.39
     bould
    1.33
     shack
    1.33
     kay
    1.28
    bah
    1.27
     canoe
    1.27
    doors
    1.25
    Activations Density 0.053%

    No Known Activations