INDEX

Explanations

directional instructions and terms related to movement

oai_token-act-pair · gpt-3.5-turbo

directional instructions related to movement and positioning

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

©¶æ¥µ

-0.68

pires

-0.67

 Schwe

-0.67

BuyableInstoreAndOnline

-0.65

 Feder

-0.64

pired

-0.62

 Anth

-0.61

atti

-0.59

Website

-0.59

Nation

-0.59

POSITIVE LOGITS

 versus

1.04

1.03

/-

1.02

or

1.00

/+

1.00

direction

0.96

 axis

0.96

vs

0.92

wards

0.90

ward

0.89

Activations Density 0.216%

directional instructions and terms related to movement

directional instructions related to movement and positioning

No Comments

No Known Activations