INDEX

Explanations

questions starting with "Can"

oai_token-act-pair · gpt-3.5-turbo

questions beginning with the word "Can."

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

VERTISEMENT

-0.67

edient

-0.63

direction

-0.62

eering

-0.60

 striving

-0.60

 honoring

-0.58

tube

-0.57

CPI

-0.57

 pane

-0.57

 thirds

-0.56

POSITIVE LOGITS

't

1.39

berra

1.23

adian

1.09

vas

0.99

NOT

0.96

opy

0.88

isters

0.83

ister

0.81

tera

0.80

atell

0.80

Activations Density 0.047%

questions starting with "Can"

questions beginning with the word "Can."

No Comments

No Known Activations