INDEX

Explanations

mentions of the term "gl" with numbers ranging from 9 to 10

oai_token-act-pair · gpt-3.5-turbo

references to gliding or gliders

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Ware

-0.74

brance

-0.72

ribe

-0.72

Hunt

-0.71

payer

-0.71

 Notification

-0.70

 Utilities

-0.68

PT

-0.68

eto

-0.68

acker

-0.67

POSITIVE LOGITS

gl

3.49

Gl

2.01

gl

1.92

Gl

1.75

gle

1.59

 glide

1.54

GL

1.54

 glitter

1.29

GL

1.28

 glow

1.26

Activations Density 0.010%

mentions of the term "gl" with numbers ranging from 9 to 10

references to gliding or gliders

No Comments

No Known Activations