INDEX

Explanations

numerical height measurements mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

references to measurements of height

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

vernment

-0.81

tical

-0.74

BOOK

-0.73

plet

-0.72

ctive

-0.71

uel

-0.69

ctory

-0.66

BLE

-0.65

redo

-0.65

Ã§a

-0.64

POSITIVE LOGITS

 height

1.32

 heights

1.19

 stature

0.94

height

0.87

 Height

0.85

 tallest

0.84

 elevation

0.84

 variance

0.81

 altitude

0.81

 distances

0.74

Activations Density 0.007%

numerical height measurements mentioned in the text

references to measurements of height

No Comments

No Known Activations