INDEX

Explanations

phrases related to statements or messages written on T-shirts or signs

oai_token-act-pair · gpt-3.5-turbo

the word "that" 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 @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

soDeliveryDate

-0.77

Plex

-0.75

 ILCS

-0.74

eworks

-0.67

amping

-0.66

antine

-0.63

*/

-0.62

 srfAttach

-0.61

OSE

-0.61

ongo

-0.60

POSITIVE LOGITS

 reads

1.17

 read

1.17

 satir

1.13

 depicts

1.10

 stated

1.10

 depicted

1.09

 purported

1.04

 contained

1.03

 mentions

1.01

 quoted

0.99

Activations Density 0.242%

phrases related to statements or messages written on T-shirts or signs

the word "that" in various contexts

No Comments

No Known Activations