© 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. 0-RES-JB
    11. 11583
    Prev
    Next
    INDEX
    Explanations

    the word "buddy" either with a very strong (value of 10) or strong (value of 9) activation

    oai_token-act-pair · gpt-3.5-turbo

    references to friendship and terms related to friends

    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
     NEC
    -0.72
    cul
    -0.70
    ruct
    -0.70
     Circ
    -0.69
    oeuv
    -0.68
     mould
    -0.68
    onduct
    -0.68
     adjourn
    -0.67
     Anglic
    -0.66
     labour
    -0.66
    POSITIVE LOGITS
     buddy
    3.07
     buddies
    2.49
     Buddy
    2.35
    uddy
    2.09
     pals
    1.28
     cozy
    1.24
     sucker
    1.22
     fuzzy
    1.17
     Jarrett
    1.08
     dude
    1.06
    Activations Density 0.033%

    No Known Activations