© 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
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 15601
    Prev
    Next
    INDEX
    Explanations

    proper nouns consisting of people's names or acronyms

    oai_token-act-pair · gpt-3.5-turbo

    the letter 'R' in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     wonders
    -0.70
     tamp
    -0.63
     cement
    -0.62
     sometime
    -0.61
     spo
    -0.59
     kindred
    -0.59
     candle
    -0.58
     moot
    -0.58
     emb
    -0.57
     catering
    -0.57
    POSITIVE LOGITS
    R
    3.33
    Rs
    2.45
     R
    1.76
    RR
    1.75
    RG
    1.68
    RH
    1.63
    Ry
    1.63
    RF
    1.59
    RM
    1.58
    r
    1.58
    Activations Density 0.028%

    No Known Activations