INDEX

Explanations

proper nouns ending in 'Jr.' with a high level of activation

oai_token-act-pair · gpt-3.5-turbo

occurrences of names followed by "Jr."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 rule

-0.66

 predator

-0.63

 snail

-0.63

 calendar

-0.62

 transact

-0.61

 cradle

-0.60

 privileged

-0.60

 hunter

-0.58

 calorie

-0.57

 dams

-0.57

POSITIVE LOGITS

atown

0.70

 sidx

0.70

 Released

0.69

 Speedway

0.68

 Macedonia

0.66

Hue

0.62

Laz

0.61

aiden

0.61

Jr

0.61

Ã±a

0.60

Activations Density 0.056%

proper nouns ending in 'Jr.' with a high level of activation

occurrences of names followed by "Jr."

No Comments

No Known Activations