INDEX

Explanations

the word "that" at a high activation level

oai_token-act-pair · gpt-3.5-turbo

the word "that" used 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 @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

EMBER

-0.66

cept

-0.64

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.63

arest

-0.62

Tank

-0.62

izont

-0.62

ãĥ¡

-0.62

Pont

-0.61

ãĥīãĥ©

-0.59

ãĤ·

-0.59

POSITIVE LOGITS

 although

0.95

"[

0.86

'[

0.76

 sounded

0.72

 whilst

0.69

soever

0.68

 while

0.67

 "...

0.66

utenberg

0.64

 contradicts

0.64

Activations Density 0.193%

the word "that" at a high activation level

the word "that" used in various contexts

No Comments

No Known Activations