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. 21059
    Prev
    Next
    INDEX
    Explanations

    mentions of words containing the ending "arent."

    oai_token-act-pair · gpt-3.5-turbo

    terms related to appearances and characteristics associated with the "Barents Sea."

    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
     pulp
    -0.76
     Kab
    -0.67
     gh
    -0.66
     Peel
    -0.66
     pain
    -0.65
     Asylum
    -0.65
     Coil
    -0.64
     tone
    -0.64
     tones
    -0.63
     Swamp
    -0.63
    POSITIVE LOGITS
    arent
    4.77
    earance
    1.74
    earances
    1.51
    arently
    1.49
    antry
    1.20
    reciation
    1.17
    rox
    1.09
    enhagen
    0.99
    lihood
    0.90
    amily
    0.88
    Activations Density 0.046%

    No Known Activations