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. 23876
    Prev
    Next
    INDEX
    Explanations

    words related to positivity and enjoyment

    oai_token-act-pair · gpt-3.5-turbo

    words associated with positivity and beauty

    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
    ache
    -0.70
    ennis
    -0.67
    ACC
    -0.67
     vs
    -0.66
    Ops
    -0.65
    Analy
    -0.63
    ence
    -0.61
     Restrict
    -0.61
     Dre
    -0.61
    ateg
    -0.61
    POSITIVE LOGITS
     lovely
    3.10
     wonderful
    2.35
     delightful
    2.34
     gorgeous
    2.01
     fabulous
    1.94
     wonderfully
    1.92
     marvelous
    1.89
     charming
    1.81
     terrific
    1.72
     adorable
    1.72
    Activations Density 0.037%

    No Known Activations