INDEX

Explanations

dates, specifically those in the format of day followed by month followed by year

oai_token-act-pair · gpt-3.5-turbo

references to the year 199

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

orc

-0.90

aminer

-0.88

ensional

-0.81

orem

-0.74

anguage

-0.74

igent

-0.72

ellation

-0.72

arling

-0.71

ional

-0.71

heed

-0.70

POSITIVE LOGITS

th

1.06

âĸĪâĸĪ

0.93

0.87

0.86

0.84

0.82

Activations Density 0.025%

dates, specifically those in the format of day followed by month followed by year

references to the year 199

No Comments

No Known Activations