INDEX

Explanations

phrases expressing astonishment or surprise related to seeing something unique or unprecedented

oai_token-act-pair · gpt-3.5-turbo

phrases expressing the notion of unprecedented experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

•layer 9 - surprise

No Comments

Negative Logits

ŃĶ

-0.69

 resid

-0.64

nr

-0.64

fund

-0.63

emies

-0.60

apters

-0.59

nels

-0.58

thren

-0.58

 ancest

-0.58

ilst

-0.58

POSITIVE LOGITS

 anything

0.97

 anybody

0.74

 anyone

0.74

nor

0.74

him

0.71

ANY

0.69

 footage

0.68

anything

0.66

Initialized

0.65

irlf

0.65

Activations Density 0.093%

phrases expressing astonishment or surprise related to seeing something unique or unprecedented

phrases expressing the notion of unprecedented experiences

No Comments

No Known Activations