INDEX

Explanations

contractions with 'won' (short for 'will not') in the context of instructions or advice

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "won't" 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 @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

OTOS

-0.71

mens

-0.66

illin

-0.63

resa

-0.60

mans

-0.60

guns

-0.59

repre

-0.59

 giants

-0.58

angering

-0.58

men

-0.58

POSITIVE LOGITS

't

1.19

itive

0.80

tyard

0.71

now

0.70

ners

0.68

ald

0.67

cest

0.66

 decisively

0.66

nings

0.65

rar

0.64

Activations Density 0.022%

contractions with 'won' (short for 'will not') in the context of instructions or advice

instances of the phrase "won't" in various contexts

No Comments

No Known Activations