© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 2547
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "BlackBerry" and seafood-related terms like "crab."

    oai_token-act-pair · gpt-3.5-turbo

    mentions of the brand "BlackBerry" and references to crab and lobster

    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
    ãĥ©
    -0.75
     masks
    -0.74
    eno
    -0.68
     Nikola
    -0.68
     Paste
    -0.68
    sam
    -0.68
    ãĥİ
    -0.68
    riks
    -0.67
     ml
    -0.66
    vill
    -0.66
    POSITIVE LOGITS
     BlackBerry
    3.10
     Lob
    1.67
     crab
    1.34
     crabs
    1.33
     Roku
    1.29
    Berry
    1.27
     Cisco
    1.26
     lobster
    1.20
     Crab
    1.18
     Roc
    1.15
    Activations Density 0.034%

    No Known Activations