INDEX

Explanations

mentions of the word "jeep"

oai_token-act-pair · gpt-3.5-turbo

the word "je" in various contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

work

-0.67

onics

-0.66

urate

-0.66

under

-0.66

power

-0.64

 Shards

-0.63

 Beacon

-0.63

aster

-0.62

ignt

-0.61

coin

-0.61

POSITIVE LOGITS

je

4.03

Je

1.81

je

1.67

 Jeep

1.61

Je

1.37

boo

1.20

 vill

1.18

sne

1.12

ja

1.09

ber

1.03

Activations Density 0.011%

mentions of the word "jeep"

the word "je" in various contexts

No Comments

No Known Activations