INDEX

Explanations

instances of numeric information preceded by the word "about"

oai_token-act-pair · gpt-3.5-turbo

the word "about" used in various contexts related to time and measurements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Peaks

-0.73

Express

-0.70

BT

-0.65

hesis

-0.63

glas

-0.62

chens

-0.61

NEY

-0.61

athing

-0.57

 Dynamics

-0.57

Write

-0.56

POSITIVE LOGITS

 halfway

0.85

0.82

0.81

0.79

0.78

 eighty

0.76

aleb

0.76

 midway

0.76

0.75

Activations Density 0.064%

instances of numeric information preceded by the word "about"

the word "about" used in various contexts related to time and measurements

No Comments

No Known Activations