INDEX

Explanations

words related to tools or objects that can be used in specific situations

oai_token-act-pair · gpt-3.5-turbo

references to tools and materials

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 ÂŃ

-0.85

âĢ

-0.76

å¼

-0.74

âĦ¢:

-0.70

ãĥ©ãĥ³

-0.70

 âĢ

-0.69

cknow

-0.68

—"

-0.68

ÂŃ

-0.67

ãĥ¼ãĥ

-0.67

POSITIVE LOGITS

 alot

1.07

 haha

1.06

etc

1.06

 myself

1.00

 doesnt

0.93

 didnt

0.93

 thats

0.90

tho

0.90

ect

0.89

 beforehand

0.88

Activations Density 0.917%

words related to tools or objects that can be used in specific situations

references to tools and materials

No Comments

No Known Activations