© 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. 3-RES-JB
    11. 13375
    Prev
    Next
    INDEX
    Explanations

    names of specific countries and organizations

    oai_token-act-pair · gpt-3.5-turbo

    specific proper nouns and names of places or entities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     direction
    -0.61
    acles
    -0.61
     trans
    -0.61
     opposite
    -0.61
     disg
    -0.59
     foreseeable
    -0.57
     fertile
    -0.56
     vicinity
    -0.55
     prestigious
    -0.55
     GF
    -0.55
    POSITIVE LOGITS
    theless
    1.13
    anyahu
    0.98
    gomery
    0.93
    withstanding
    0.89
    resa
    0.89
    odore
    0.88
    agascar
    0.87
    tenance
    0.87
    bnb
    0.87
    mosp
    0.87
    Activations Density 0.297%

    No Known Activations