INDEX

Explanations

questions beginning with the word "Are."

oai_token-act-pair · gpt-3.5-turbo

questions that start with "Are" addressing various subjects or situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

oire

-0.70

ching

-0.67

ãĤ¨ãĥ«

-0.65

ð

-0.64

âĶĢâĶĢ

-0.63

ulates

-0.63

ulated

-0.62

rift

-0.62

DCS

-0.62

oting

-0.61

POSITIVE LOGITS

nt

0.97

senal

0.82

wolves

0.81

 gonna

0.77

NOT

0.75

nda

0.74

ppo

0.71

jon

0.70

ync

0.70

ethe

0.69

Activations Density 0.019%

questions beginning with the word "Are."

questions that start with "Are" addressing various subjects or situations

No Comments

No Known Activations