INDEX

Explanations

keywords or commands indicating the end or closure of a segment or document, such as "CLOSE."

oai_token-act-pair · gpt-3.5-turbo

instances of the word "CLOSE" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

hid

-0.88

 disgu

-0.82

beh

-0.77

ding

-0.77

formation

-0.76

 dece

-0.75

osc

-0.74

foreseen

-0.74

ministic

-0.74

ifice

-0.73

POSITIVE LOGITS

 Officials

1.19

CLOSE

1.10

 Hundreds

1.10

 Thousands

1.05

 Former

1.05

 Researchers

1.05

 Investigators

1.04

 Protesters

1.03

 Dozens

1.02

 Authorities

1.00

Activations Density 0.017%

keywords or commands indicating the end or closure of a segment or document, such as "CLOSE."

instances of the word "CLOSE" in various contexts

No Comments

No Known Activations