© 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
    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. 6553
    Prev
    Next
    INDEX
    Explanations

    phrases related to the concept of using people or things as tools or instruments

    oai_token-act-pair · gpt-3.5-turbo

    instances of political manipulation involving marginalized groups and their rights

    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
    dt
    -0.69
    omes
    -0.69
    eeper
    -0.66
    etheless
    -0.64
    existence
    -0.63
     arrivals
    -0.63
    uez
    -0.62
     Aware
    -0.61
    ORTS
    -0.60
    atri
    -0.58
    POSITIVE LOGITS
     leverage
    1.32
     pretext
    1.23
     excuse
    1.19
     fodder
    1.06
     blackmail
    1.04
     justification
    1.03
     bait
    1.03
    weapon
    1.02
     ammunition
    1.02
     bargaining
    1.02
    Activations Density 0.259%

    No Known Activations