INDEX

Explanations

words related to media, entertainment, and publications

oai_token-act-pair · gpt-3.5-turbo

specific video game titles and related terminology

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

perature

-0.77

 customary

-0.76

 referen

-0.74

 volunte

-0.72

 seiz

-0.72

uphem

-0.70

 Mehran

-0.69

 inexpensive

-0.69

 authentic

-0.69

 skilled

-0.67

POSITIVE LOGITS

 Studios

1.19

 Skies

1.07

 Technologies

1.06

 Labs

1.05

 Empires

1.05

 Brewing

1.04

 Entertainment

1.04

 Records

1.03

 Chronicles

1.03

 Tactics

1.03

Activations Density 0.412%

words related to media, entertainment, and publications

specific video game titles and related terminology

No Comments

No Known Activations