INDEX

Explanations

phrases related to changes in quantity or degree

oai_token-act-pair · gpt-3.5-turbo

terms that indicate increases or decreases in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DragonMagazine

-0.85

mad

-0.78

çĦ

-0.73

Sham

-0.72

oof

-0.69

ilk

-0.68

odore

-0.67

ivot

-0.65

ubi

-0.63

Mich

-0.63

POSITIVE LOGITS

 productivity

1.23

 efficiency

1.14

 visibility

1.13

 profitability

1.07

 competitiveness

1.04

 circulation

1.03

 chances

1.02

 morale

1.01

 perceptions

1.00

 overall

1.00

Activations Density 0.173%

phrases related to changes in quantity or degree

terms that indicate increases or decreases in various contexts

No Comments

No Known Activations