INDEX

Explanations

names of people or places

oai_token-act-pair · gpt-3.5-turbo

proper nouns and terms related to significant themes or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Typhoon

-0.65

à©

-0.63

Hok

-0.62

DERR

-0.62

 Amendments

-0.59

 Virt

-0.58

IRD

-0.57

 Brah

-0.57

 Crunch

-0.56

 Hispan

-0.55

POSITIVE LOGITS

 wise

1.09

matically

1.02

ilaterally

0.98

rily

0.90

ically

0.89

wise

0.88

lessly

0.86

cause

0.84

pecially

0.84

ecause

0.82

Activations Density 0.862%

names of people or places

proper nouns and terms related to significant themes or concepts

No Comments

No Known Activations