INDEX

Explanations

phrases related to uncertainty or speculation

oai_token-act-pair · gpt-3.5-turbo

phrases that express uncertainty or questions about knowledge and opinions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 srfAttach

-0.58

bryce

-0.55

catentry

-0.54

 respectively

-0.52

anwhile

-0.50

accompanied

-0.50

Pwr

-0.49

 runoff

-0.47

 refill

-0.47

ornia

-0.46

POSITIVE LOGITS

 miscar

0.56

 spoilers

0.53

 nodd

0.53

_>

0.52

 explanations

0.52

 motives

0.50

 mistakes

0.50

vez

0.49

 endings

0.47

;)

0.46

Activations Density 2.179%

phrases related to uncertainty or speculation

phrases that express uncertainty or questions about knowledge and opinions

No Comments

No Known Activations