INDEX

Explanations

information regarding quantities or numerical values

oai_token-act-pair · gpt-3.5-turbo

comparative phrases that quantify amounts exceeding a specific threshold

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ACTED

-0.67

 largeDownload

-0.65

 Finder

-0.65

NEY

-0.63

agate

-0.63

 derog

-0.62

 Bomber

-0.61

OSP

-0.60

inventoryQuantity

-0.60

ãĤ«

-0.60

POSITIVE LOGITS

 dozen

0.78

whelming

0.77

vernment

0.74

dozen

0.74

represent

0.74

lyak

0.73

0.71

0.70

capacity

0.68

 hundred

0.67

Activations Density 0.049%

information regarding quantities or numerical values

comparative phrases that quantify amounts exceeding a specific threshold

No Comments

No Known Activations