INDEX

Explanations

phrases indicating proximity or location

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate a specific location or moment in time

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

amar

-0.75

crew

-0.74

anship

-0.73

imens

-0.72

gins

-0.71

imb

-0.71

wcs

-0.70

orously

-0.70

iple

-0.69

ply

-0.68

POSITIVE LOGITS

Zur

0.65

enegger

0.64

 Razor

0.64

 Rampage

0.63

 Aberdeen

0.63

 doorstep

0.62

 Tooth

0.61

 Stockholm

0.60

çĶ°

0.60

ol

0.60

Activations Density 0.110%

phrases indicating proximity or location

phrases that indicate a specific location or moment in time

No Comments

No Known Activations