INDEX

Explanations

terms related to a specific brand or concept, "Shinobi"

oai_token-act-pair · gpt-3.5-turbo

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

ivities

-0.78

 ORIG

-0.68

 cort

-0.67

 Orient

-0.65

imates

-0.64

 Paran

-0.64

 butterflies

-0.63

 Quan

-0.62

 waters

-0.61

 oceans

-0.60

POSITIVE LOGITS

server

1.25

edient

1.25

edience

1.22

acter

1.21

esity

1.12

ranch

1.10

fusc

1.10

served

1.09

livion

1.08

acco

1.07

Activations Density 0.023%

terms related to a specific brand or concept, "Shinobi"

No Comments

No Known Activations