INDEX

Explanations

phrases related to scrolling for additional content or videos

oai_token-act-pair · gpt-3.5-turbo

instructions or prompts to scroll within content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

stood

-0.83

roman

-0.72

inventoryQuantity

-0.70

ãĤº

-0.63

rule

-0.61

determination

-0.61

 partnership

-0.60

 consequ

-0.60

chal

-0.59

 bonding

-0.58

POSITIVE LOGITS

Scroll

1.02

 Scroll

1.00

scroll

0.80

ificant

0.75

tip

0.74

Below

0.74

ogue

0.72

Recipe

0.72

 scroll

0.71

Coin

0.71

Activations Density 0.007%

phrases related to scrolling for additional content or videos

instructions or prompts to scroll within content

No Comments

No Known Activations