INDEX

Explanations

terms related to elimination or removal

oai_token-act-pair · gpt-3.5-turbo

terms related to elimination processes

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

atern

-0.68

acca

-0.68

 Furious

-0.67

Tah

-0.65

 Indust

-0.65

lying

-0.65

Pis

-0.65

ebook

-0.63

rain

-0.63

 Arbor

-0.62

POSITIVE LOGITS

 elimination

1.32

 elim

1.09

itaire

1.02

ptive

0.92

 Elim

0.88

 eliminated

0.87

ptives

0.84

inators

0.81

culosis

0.81

ission

0.78

Activations Density 0.006%

terms related to elimination or removal

terms related to elimination processes

No Comments

No Known Activations