INDEX

Explanations

descriptions of events

oai_token-act-pair · gpt-3.5-turbo

references to significant occurrences or situations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

zzi

-0.76

tesy

-0.72

geon

-0.67

 porous

-0.67

 territorial

-0.66

sis

-0.66

Maz

-0.65

 Sound

-0.64

leted

-0.63

zie

-0.63

POSITIVE LOGITS

 events

1.36

 Events

1.02

events

0.99

 event

0.95

 unfolded

0.92

 transpired

0.89

hips

0.87

 unfold

0.87

hower

0.86

 festivities

0.85

Activations Density 0.017%

descriptions of events

references to significant occurrences or situations

No Comments

No Known Activations