INDEX

Explanations

phrases related to specific medical terms or scientific concepts

oai_token-act-pair · gpt-3.5-turbo

the word "aph" in various contexts, likely related to specific terminology

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

PUT

-0.82

âĸ¬âĸ¬

-0.75

swick

-0.72

 Line

-0.68

NING

-0.66

 Borderlands

-0.65

 Supplemental

-0.65

present

-0.64

 LINE

-0.63

ÂłÂłÂłÂł

-0.62

POSITIVE LOGITS

aph

1.31

obia

1.17

rod

1.01

onso

0.99

ysics

0.98

alon

0.98

obic

0.97

amily

0.96

ornia

0.96

osate

0.96

Activations Density 0.004%

phrases related to specific medical terms or scientific concepts

the word "aph" in various contexts, likely related to specific terminology

No Comments

No Known Activations