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. 8815
    Prev
    Next
    INDEX
    Explanations

    names containing the string "acey" or "Alex" followed by a number

    oai_token-act-pair · gpt-3.5-turbo

    proper nouns, particularly personal names

    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
     ibn
    -0.71
     Centauri
    -0.68
     wid
    -0.67
     Nobel
    -0.66
     Nib
    -0.66
     hamb
    -0.65
     ponder
    -0.64
     barrel
    -0.64
     Hum
    -0.63
     nib
    -0.63
    POSITIVE LOGITS
    acey
    2.97
    Alex
    1.62
    eny
    1.52
    acy
    1.32
     Alexis
    1.16
     Alexandria
    1.10
     Alexa
    1.06
    ilee
    1.03
     Lauren
    1.02
    icia
    0.99
    Activations Density 0.032%

    No Known Activations