INDEX

Explanations

words related to easing, relieving, alleviating, or mitigating various issues or problems

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate the reduction of barriers or difficulties

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

ertodd

-0.73

ittal

-0.72

ymes

-0.71

aries

-0.70

 Kinnikuman

-0.67

 Mahjong

-0.63

uese

-0.63

prints

-0.63

acular

-0.62

ograms

-0.62

POSITIVE LOGITS

 tensions

1.31

 boredom

1.25

 frustrations

1.15

 congestion

1.13

 tension

1.12

 overcrowd

1.09

 discomfort

1.07

 burdens

1.06

 burden

1.06

 fears

1.05

Activations Density 0.118%

words related to easing, relieving, alleviating, or mitigating various issues or problems

phrases that indicate the reduction of barriers or difficulties

No Comments

No Known Activations