INDEX

Explanations

marketing-related terms like "updates" and "promotions"

oai_token-act-pair · gpt-3.5-turbo

terms related to updates and promotions from a news source

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

conn

-0.59

efer

-0.53

 denomin

-0.50

oun

-0.50

asus

-0.50

 Secondly

-0.50

amina

-0.50

hog

-0.50

SAN

-0.50

abase

-0.49

POSITIVE LOGITS

interstitial

0.67

 categorized

0.60

 widest

0.56

felt

0.56

terms

0.55

piracy

0.55

 promotions

0.55

auri

0.54

Afee

0.54

 Featured

0.54

Activations Density 0.030%

marketing-related terms like "updates" and "promotions"

terms related to updates and promotions from a news source

No Comments

No Known Activations