INDEX

Explanations

phrases indicating events or announcements that will happen in the near future

oai_token-act-pair · gpt-3.5-turbo

references to time-related phrases

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ulner

-0.79

cu

-0.73

ward

-0.72

WARD

-0.71

antz

-0.71

omas

-0.70

uates

-0.70

ohn

-0.70

agascar

-0.69

acter

-0.69

POSITIVE LOGITS

 meantime

1.47

 vein

1.13

 vicinity

1.06

 evenings

1.01

 midst

0.99

 hopes

0.93

 mornings

0.92

 Philippines

0.92

 afternoon

0.91

 meanwhile

0.90

Activations Density 0.179%

phrases indicating events or announcements that will happen in the near future

references to time-related phrases

No Comments

No Known Activations