INDEX

Explanations

mentions of the word "bee"

oai_token-act-pair · gpt-3.5-turbo

references to "bee" or "bees" in the text

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

ional

-0.87

igators

-0.77

aries

-0.75

igator

-0.74

ories

-0.74

 Honour

-0.74

 Balkans

-0.72

ablishment

-0.69

icked

-0.66

 Archdemon

-0.65

POSITIVE LOGITS

bee

1.56

pee

1.10

Bee

1.07

bees

1.06

Bee

1.06

bee

1.03

ffe

1.02

zeb

0.96

hyde

0.93

gee

0.92

Activations Density 0.006%

mentions of the word "bee"

references to "bee" or "bees" in the text

No Comments

No Known Activations