INDEX

Explanations

words related to "CT," which likely stands for a specific term or entity

oai_token-act-pair · gpt-3.5-turbo

references to Connecticut (CT)

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

wagen

-0.87

volent

-0.75

leased

-0.73

bye

-0.70

morph

-0.70

 burgers

-0.69

itably

-0.69

conservancy

-0.69

olulu

-0.68

lihood

-0.67

POSITIVE LOGITS

RL

1.07

ARDIS

0.96

CT

0.95

ICLE

0.89

JM

0.84

arre

0.83

ARGET

0.83

IA

0.80

KO

0.80

ombs

0.80

Activations Density 0.007%

words related to "CT," which likely stands for a specific term or entity

references to Connecticut (CT)

No Comments

No Known Activations