INDEX

Explanations

phrases related to progress or advancement

oai_token-act-pair · gpt-3.5-turbo

references to progress or advancements measured in steps

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

ecause

-0.82

liction

-0.77

eer

-0.72

nces

-0.70

elta

-0.67

licts

-0.67

iqueness

-0.65

inately

-0.63

licted

-0.63

egu

-0.63

POSITIVE LOGITS

 forward

1.29

 backward

1.18

 backwards

1.11

 forwards

1.09

 toward

1.07

 towards

1.04

 closer

0.99

 Forward

0.99

frog

0.99

forward

0.96

Activations Density 0.023%

phrases related to progress or advancement

references to progress or advancements measured in steps

No Comments

No Known Activations