INDEX

Explanations

news agency names

oai_token-act-pair · gpt-3.5-turbo

repeated references to the news organization Reuters

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

izers

-0.74

Ö¼

-0.66

ivas

-0.65

ysis

-0.64

ount

-0.64

gran

-0.63

 Dynam

-0.60

uster

-0.60

izer

-0.60

izable

-0.59

POSITIVE LOGITS

UTERS

0.97

 REUTERS

0.95

CLASSIFIED

0.85

externalActionCode

0.84

Film

0.83

WAR

0.80

photo

0.78

WARD

0.78

FORE

0.78

AMI

0.77

Activations Density 0.008%

news agency names

repeated references to the news organization Reuters

No Comments

No Known Activations