INDEX

Explanations

phrases indicating a place or environment that elicits a strong emotional response

oai_token-act-pair · gpt-3.5-turbo

phrases indicating places or situations that are enjoyable or undesirable

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Provided

-0.72

 Limits

-0.69

afety

-0.68

otti

-0.67

sheets

-0.66

arity

-0.66

oret

-0.66

approved

-0.64

vich

-0.64

limits

-0.61

POSITIVE LOGITS

 contemplate

1.19

 begin

1.16

 start

1.11

 introduce

1.05

 watch

1.01

 invest

1.01

 undertake

1.01

 acquire

1.00

 spend

0.98

 cultivate

0.98

Activations Density 0.108%

phrases indicating a place or environment that elicits a strong emotional response

phrases indicating places or situations that are enjoyable or undesirable

No Comments

No Known Activations