INDEX

Explanations

phrases related to specific names or locations mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

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

 Internal

-0.76

 Tent

-0.73

 Merrill

-0.71

 Method

-0.71

 Mutual

-0.71

Internal

-0.70

 exchanging

-0.70

Ill

-0.68

CLASSIFIED

-0.68

 Nicarag

-0.68

POSITIVE LOGITS

ham

1.09

achy

1.05

Ham

1.01

HAM

1.01

agh

1.00

achus

0.95

rag

0.90

ras

0.88

rat

0.88

Ham

0.88

Activations Density 0.301%

phrases related to specific names or locations mentioned in the text

No Comments

No Known Activations