© 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. 2799
    Prev
    Next
    INDEX
    Explanations

    character actions and revelations within a storyline

    oai_token-act-pair · gpt-3.5-turbo

    key plot developments and character discoveries in a narrative

    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
    ongyang
    -1.09
    itutional
    -0.85
    ijing
    -0.85
    espie
    -0.84
    conom
    -0.80
    arij
    -0.78
    ividual
    -0.76
    hower
    -0.76
    ricular
    -0.75
    velop
    -0.73
    POSITIVE LOGITS
     flashbacks
    1.00
     flashback
    0.98
     herself
    0.95
     Joker
    0.92
     Valkyrie
    0.90
     Kenobi
    0.90
     Detective
    0.89
     himself
    0.85
     Loki
    0.84
     Ichigo
    0.83
    Activations Density 1.297%

    No Known Activations