INDEX

Explanations

words associated with negative conditions or outcomes

oai_token-act-pair · gpt-3.5-turbo

phrases related to negative evaluations or descriptions of situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 shutter

-0.65

thouse

-0.64

zees

-0.62

 Rouge

-0.61

 crumble

-0.61

 Sharing

-0.60

 lots

-0.60

 rapists

-0.58

 Closed

-0.58

 Delete

-0.58

POSITIVE LOGITS

gotten

1.34

fitting

1.22

treatment

1.19

fortune

1.12

intent

1.06

founded

1.06

informed

1.05

defined

1.05

effects

1.02

equipped

1.02

Activations Density 0.043%

words associated with negative conditions or outcomes

phrases related to negative evaluations or descriptions of situations

No Comments

No Known Activations