INDEX

Explanations

mentions of a specific location (Arlington) in different contexts

oai_token-act-pair · gpt-3.5-turbo

mentions of "Arlington" and related locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ocide

-0.77

ggle

-0.73

udding

-0.71

arist

-0.70

icism

-0.69

osexual

-0.69

onym

-0.69

udic

-0.68

oly

-0.67

cribed

-0.67

POSITIVE LOGITS

lington

0.97

INGTON

0.90

 Heights

0.85

pole

0.85

 Worth

0.85

 Mavericks

0.75

 County

0.75

IED

0.70

 Metropolitan

0.70

 Kendall

0.69

Activations Density 0.018%

mentions of a specific location (Arlington) in different contexts

mentions of "Arlington" and related locations

No Comments

No Known Activations