INDEX

Explanations

phrases related to taking action and making decisions

oai_token-act-pair · gpt-3.5-turbo

repeated references to the word "we."

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

REDACTED

-0.72

 Publication

-0.71

 gratification

-0.67

odor

-0.66

ions

-0.62

Tai

-0.59

Nay

-0.58

more

-0.58

 Eleven

-0.58

 Rowe

-0.57

POSITIVE LOGITS

've

1.32

're

1.27

'll

1.09

asel

1.07

 ourselves

1.07

'd

1.05

athered

1.03

IRD

1.02

ibo

0.95

lder

0.94

Activations Density 0.258%

phrases related to taking action and making decisions

repeated references to the word "we."

No Comments

No Known Activations