© 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
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 11-RES-JB
    11. 841
    Prev
    Next
    INDEX
    Explanations

    phrases indicating strong recommendation or necessity

    oai_token-act-pair · gpt-3.5-turbo

    suggestions for essential items or experiences

    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
    ufact
    -0.72
    quickShipAvailable
    -0.61
    itri
    -0.58
     lull
    -0.58
     Zo
    -0.56
    etically
    -0.54
    orp
    -0.53
     terminating
    -0.53
     multiplying
    -0.52
     Emanuel
    -0.52
    POSITIVE LOGITS
    have
    1.05
    watch
    1.00
    reads
    0.97
    dos
    0.96
    read
    0.95
    list
    0.94
     see
    0.92
     READ
    0.92
    see
    0.90
    buy
    0.89
    Activations Density 0.068%

    No Known Activations