INDEX

Explanations

statements indicating possibility or capability

oai_token-act-pair · gpt-3.5-turbo

instances of the word "can" 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

quartered

-0.73

ONSORED

-0.68

 referen

-0.65

interstitial

-0.63

 Rutherford

-0.60

VICE

-0.59

ASED

-0.59

Lens

-0.59

 Seeking

-0.59

 Likes

-0.59

POSITIVE LOGITS

't

2.07

 afford

1.27

NOT

1.15

berra

1.08

adian

1.06

 attest

0.99

 easily

0.99

 handle

0.97

 hardly

0.97

isters

0.96

Activations Density 0.162%

statements indicating possibility or capability

instances of the word "can" in various contexts

No Comments

No Known Activations