INDEX

Explanations

verbs or phrases related to changing one's mind or opinion

oai_token-act-pair · gpt-3.5-turbo

references to changes in opinion or stance

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

ocument

-0.83

 Awakens

-0.81

ĳå£«

-0.76

etsk

-0.75

ulhu

-0.74

cade

-0.74

urat

-0.74

aphael

-0.74

isson

-0.73

lis

-0.71

POSITIVE LOGITS

 tune

1.41

 mind

1.41

 minds

1.33

 mindset

1.18

 stance

1.17

 attitude

1.06

 tack

1.06

 allegiance

1.05

 ways

1.04

 outlook

1.04

Activations Density 0.074%

verbs or phrases related to changing one's mind or opinion

references to changes in opinion or stance

No Comments

No Known Activations