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. 789
    Prev
    Next
    INDEX
    Explanations

    Unicode characters that are specific to a particular language

    oai_token-act-pair · gpt-3.5-turbo

    unusual or non-standard characters and sequences

    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
    atche
    -0.86
    20439
    -0.74
     Franch
    -0.72
     Wynne
    -0.72
     Pearce
    -0.70
     Forth
    -0.70
     Scythe
    -0.70
    orkshire
    -0.69
    ttes
    -0.69
    toe
    -0.68
    POSITIVE LOGITS
    §
    1.19
    ĥ
    1.16
    ĭ
    1.13
    ŀ
    1.11
    ķ
    1.11
    ļ
    1.09
    Ĭ
    1.05
    ĺ
    1.04
    ł
    1.04
    Ĥ¬
    1.02
    Activations Density 0.003%

    No Known Activations