INDEX

Explanations

text related to exceeding limits or boundaries

oai_token-act-pair · gpt-3.5-turbo

phrases related to limits or boundaries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

drug

-0.62

netflix

-0.60

 Everyday

-0.60

Elect

-0.60

Japanese

-0.59

Pak

-0.59

HA

-0.59

xiety

-0.59

online

-0.58

mart

-0.58

POSITIVE LOGITS

 bounds

4.18

 limits

1.38

 confines

1.38

 boundaries

1.34

 bounded

1.24

 Limits

1.20

limits

1.18

 bound

1.16

 constraints

1.06

 restraints

1.05

Activations Density 0.010%

text related to exceeding limits or boundaries

phrases related to limits or boundaries

No Comments

No Known Activations