INDEX

Explanations

mentions of a specific location named "our"

oai_token-act-pair · gpt-3.5-turbo

the word "our."

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

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

 steroids

-0.69

ulator

-0.67

aeda

-0.65

ulators

-0.64

 Robo

-0.62

 buckle

-0.62

 release

-0.60

 Rasmussen

-0.59

 steroid

-0.58

 electronically

-0.57

POSITIVE LOGITS

dain

1.44

neau

1.34

neys

1.12

dan

1.04

ishment

0.99

tesy

0.96

din

0.95

nee

0.92

inho

0.92

ished

0.91

Activations Density 0.026%

mentions of a specific location named "our"

the word "our."

No Comments

No Known Activations