Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 2-RES-JB
    6. 17280
    Prev
    Next
    INDEX
    Explanations

    proper nouns and specific names, such as "Pratt" and "Reno."

    oai_token-act-pair · gpt-3.5-turbo

    references to specific characters or elements associated with a particular narrative or theme

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    emies
    -0.68
     Heller
    -0.64
     Continent
    -0.59
     Freak
    -0.59
    POL
    -0.59
     Atlantis
    -0.59
    ©¶æ
    -0.58
     Plato
    -0.57
     Tamil
    -0.57
     GOODMAN
    -0.56
    POSITIVE LOGITS
    ruff
    1.23
    shaw
    0.90
    puff
    0.87
    street
    0.83
    dot
    0.78
    stra
    0.78
    essional
    0.78
    strap
    0.78
    inances
    0.78
    qv
    0.78
    Activations Density 0.016%

    No Known Activations