INDEX

Explanations

pronouns related to future actions or aspirations

oai_token-act-pair · gpt-3.5-turbo

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

ãĤ»

-0.65

ĸļ

-0.60

iameter

-0.58

 reminder

-0.57

 puzzling

-0.56

 Rumble

-0.55

emphasis

-0.55

ĨĴ

-0.55

 unexplained

-0.54

 questioning

-0.54

POSITIVE LOGITS

 will

1.45

'll

1.41

will

1.29

 WILL

1.20

 would

1.16

would

1.11

 someday

1.09

 could

1.07

 might

1.06

 wont

1.06

Activations Density 0.336%

pronouns related to future actions or aspirations

No Comments

No Known Activations