INDEX

Explanations

references to shoppers

oai_token-act-pair · gpt-3.5-turbo

terms related to consumers or participants in various activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 curing

-0.67

 Quit

-0.66

 limitation

-0.65

irc

-0.64

amation

-0.64

mentioned

-0.64

 fixing

-0.63

olutions

-0.63

 naming

-0.62

ortium

-0.61

POSITIVE LOGITS

 shoppers

3.15

goers

1.99

 commuters

1.79

 motorists

1.67

 pedestrians

1.46

 drinkers

1.38

 revel

1.35

 onlook

1.35

 attendees

1.31

 viewers

1.26

Activations Density 0.032%

references to shoppers

terms related to consumers or participants in various activities

No Comments

No Known Activations