INDEX

Explanations

proper nouns related to specific locations (states, cities, etc.)

oai_token-act-pair · gpt-3.5-turbo

specific legal and geographical references in the text

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

rencies

-0.87

assies

-0.78

urches

-0.74

nels

-0.73

ernels

-0.72

brates

-0.71

redients

-0.71

ctors

-0.70

anwhile

-0.69

lees

-0.68

POSITIVE LOGITS

 staffer

1.00

 colleague

0.90

 representative

0.86

 spokesperson

0.80

 variant

0.78

 exclusive

0.76

 lawyer

0.75

 delegation

0.74

 advisory

0.74

 employee

0.73

Activations Density 0.338%

proper nouns related to specific locations (states, cities, etc.)

specific legal and geographical references in the text

No Comments

No Known Activations