© 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. 11-RES-JB
    6. 9986
    Prev
    Next
    INDEX
    Explanations

    specific objects or traits that are unique or provide enhancements within a given context

    oai_token-act-pair · gpt-3.5-turbo

    sentences containing relative pronouns or conjunctions that introduce clauses

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     fearing
    -0.75
     hating
    -0.70
     bruising
    -0.65
    broken
    -0.63
    esi
    -0.63
     starved
    -0.62
     fears
    -0.62
    heard
    -0.62
     believing
    -0.60
     panicked
    -0.59
    POSITIVE LOGITS
     comprise
    1.35
     relate
    1.34
     include
    1.28
     are
    1.28
     vary
    1.25
     differ
    1.23
     accompany
    1.22
     enhance
    1.21
     represent
    1.20
     contain
    1.19
    Activations Density 0.204%

    No Known Activations