INDEX

Explanations

phrases related to disparities or differences between two entities

oai_token-act-pair · gpt-3.5-turbo

references to "gap" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

oran

-0.77

ise

-0.68

ocations

-0.64

uxe

-0.62

abad

-0.61

ovy

-0.61

AND

-0.60

asting

-0.60

odder

-0.58

gg

-0.58

POSITIVE LOGITS

gap

1.00

wid

0.94

Gap

0.90

 gaps

0.88

 between

0.86

 brid

0.86

 widened

0.84

 junction

0.80

hole

0.77

between

0.76

Activations Density 0.019%

phrases related to disparities or differences between two entities

references to "gap" in various contexts

No Comments

No Known Activations