INDEX

Explanations

the word "Low" with a high activation value

oai_token-act-pair · gpt-3.5-turbo

references to specific geographical or political entities, particularly those associated with the term "Low."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

iosyncr

-0.83

icable

-0.83

oused

-0.80

äºĶ

-0.75

PsyNetMessage

-0.71

ascular

-0.70

IBLE

-0.69

 photoc

-0.68

ailed

-0.66

 externalToEVAOnly

-0.66

POSITIVE LOGITS

 Definition

0.92

 Term

0.88

 Source

0.87

 Dealer

0.86

 Ones

0.85

 Thing

0.85

 Countries

0.84

erton

0.83

 Guys

0.83

 Investigations

0.83

Activations Density 0.123%

the word "Low" with a high activation value

references to specific geographical or political entities, particularly those associated with the term "Low."

No Comments

No Known Activations