INDEX

Explanations

mentions of or references to "The Washington Post."

oai_token-act-pair · gpt-3.5-turbo

references to the "Washington Post" and other news publications

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

',"

-0.66

wana

-0.61

figure

-0.60

ãĤ¯

-0.58

stood

-0.58

roma

-0.57

,'"

-0.57

adobe

-0.56

lift

-0.56

ologically

-0.55

POSITIVE LOGITS

 IMAGES

0.80

Buy

0.78

 FILE

0.78

via

0.74

 NETWORK

0.72

 Wire

0.69

Buy

0.67

 Staff

0.65

 CONTR

0.63

 COMM

0.62

Activations Density 0.040%

mentions of or references to "The Washington Post."

references to the "Washington Post" and other news publications

No Comments

No Known Activations