INDEX

Explanations

phrases that indicate a revelation or realization

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 @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

aukee

-0.70

raq

-0.63

gur

-0.63

orah

-0.63

waters

-0.63

stead

-0.62

ield

-0.60

ien

-0.60

thro

-0.57

tails

-0.57

POSITIVE LOGITS

 pesky

0.84

 they

0.80

 fateful

0.79

cher

0.77

 THEY

0.71

>>>>

0.66

 there

0.66

chy

0.66

izoph

0.65

 although

0.64

Activations Density 0.297%

phrases that indicate a revelation or realization

the word "that" in various contexts

No Comments

No Known Activations