INDEX

Explanations

specific information about different categories or entities, focusing on comparisons or rankings

oai_token-act-pair · gpt-3.5-turbo

repeated instances of the word "each."

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

bows

-0.73

stones

-0.73

ahs

-0.72

mares

-0.72

children

-0.70

haven

-0.70

models

-0.69

boys

-0.69

ELF

-0.69

lights

-0.68

POSITIVE LOGITS

 successive

1.30

 respective

1.11

 iteration

1.10

 individual

1.05

 element

0.95

 succeeding

0.93

 dimension

0.92

 participant

0.91

 ingredient

0.90

 facet

0.88

Activations Density 0.055%

specific information about different categories or entities, focusing on comparisons or rankings

repeated instances of the word "each."

No Comments

No Known Activations