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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 8-RES-JB
    6. 20090
    Prev
    Next
    INDEX
    Explanations

    C++ programming language related syntax and structures, including elements like iterators, standard library functions, and preprocessor directives

    oai_token-act-pair · gpt-3.5-turbo

    references to standard library components and iterator types in programming

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    lain
    -0.73
     Peninsula
    -0.70
    apy
    -0.68
    arding
    -0.65
    ysis
    -0.64
     surn
    -0.63
    edin
    -0.63
     moons
    -0.62
     peninsula
    -0.61
    jong
    -0.60
    POSITIVE LOGITS
    oren
    0.81
    zeb
    0.79
    ¡
    0.77
    endium
    0.77
    icial
    0.76
    Ĥ
    0.75
    ¦
    0.75
    ¼
    0.75
    ī
    0.74
    omatic
    0.74
    Activations Density 0.091%

    No Known Activations