INDEX

Explanations

comparative expressions using the word "than"

oai_token-act-pair · gpt-3.5-turbo

comparative phrases that highlight superiority or inferiority

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 centr

-0.67

maid

-0.67

 stead

-0.64

ALE

-0.61

keeper

-0.61

 Purg

-0.59

 largeDownload

-0.58

 clearance

-0.58

roller

-0.58

 bearer

-0.58

POSITIVE LOGITS

atos

1.17

than

0.99

ormal

0.90

vis

0.84

stellar

0.76

lihood

0.71

iors

0.71

izoph

0.71

average

0.70

thur

0.70

Activations Density 0.011%

comparative expressions using the word "than"

comparative phrases that highlight superiority or inferiority

No Comments

No Known Activations