INDEX

Explanations

mentions of professionals, specifically chefs

oai_token-act-pair · gpt-3.5-turbo

references to skilled professionals and their attitudes or actions in various contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Territories

-0.90

 Beaver

-0.68

Operation

-0.65

 Verb

-0.62

 Coast

-0.61

 Railroad

-0.59

 Operation

-0.59

 Rover

-0.59

 Maurit

-0.58

 Wonderland

-0.58

POSITIVE LOGITS

hip

1.33

paces

1.23

ervatives

1.19

hips

1.11

mith

1.09

ervative

0.95

ynthesis

0.92

pace

0.91

folk

0.91

aurus

0.90

Activations Density 0.249%

mentions of professionals, specifically chefs

references to skilled professionals and their attitudes or actions in various contexts

No Comments

No Known Activations