INDEX

Explanations

percentages or numerical values mentioned in sentences

oai_token-act-pair · gpt-3.5-turbo

numerical data related to statistics or measurements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

anian

-0.66

 lawy

-0.65

ezvous

-0.64

lay

-0.60

gets

-0.60

lying

-0.58

folk

-0.57

hire

-0.57

lings

-0.56

 tension

-0.56

POSITIVE LOGITS

1.26

 percent

1.14

%-

1.12

%:

1.10

%"

1.07

 ILCS

1.06

%,

1.04

%;

1.03

percent

1.02

mph

1.00

Activations Density 0.148%

percentages or numerical values mentioned in sentences

numerical data related to statistics or measurements

No Comments

No Known Activations