INDEX

Explanations

location-based information related to organizations or individuals

oai_token-act-pair · gpt-3.5-turbo

phrases indicating geographic locations or origins

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ormonal

-0.69

 pedals

-0.68

hetical

-0.66

 dexterity

-0.64

amsung

-0.63

 accompl

-0.62

blance

-0.60

ctive

-0.60

++;

-0.60

Applic

-0.60

POSITIVE LOGITS

 Toll

0.81

ĸļ

0.77

 suburbs

0.75

 borough

0.75

 suburb

0.73

waters

0.73

 Ahead

0.72

 Airport

0.71

quartered

0.71

 Centre

0.71

Activations Density 0.563%

location-based information related to organizations or individuals

phrases indicating geographic locations or origins

No Comments

No Known Activations