INDEX

Explanations

phrases related to rankings or lists, specifically mentioning the "top" of something

oai_token-act-pair · gpt-3.5-turbo

references to rankings or positions within a specified group

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ajor

-0.77

Mub

-0.67

[|

-0.60

ripp

-0.59

ary

-0.58

 Lauder

-0.58

rain

-0.55

 NOTICE

-0.55

thens

-0.55

 Daylight

-0.54

POSITIVE LOGITS

most

1.10

ography

1.03

ographical

1.03

eka

0.97

 tier

0.96

 earners

0.90

ographic

0.89

 contenders

0.89

 notch

0.88

 scorer

0.86

Activations Density 0.037%

phrases related to rankings or lists, specifically mentioning the "top" of something

references to rankings or positions within a specified group

No Comments

No Known Activations