INDEX

Explanations

references to locations inside a vehicle, specifically focusing on the back seat

oai_token-act-pair · gpt-3.5-turbo

references to the locations of individuals within vehicles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

upon

-0.81

_-

-0.75

andra

-0.75

milo

-0.74

avorite

-0.68

ANI

-0.68

DCS

-0.67

 Parenthood

-0.65

edIn

-0.65

inventoryQuantity

-0.65

POSITIVE LOGITS

seat

1.08

 seat

1.02

woods

0.96

 door

0.96

 porch

0.92

 window

0.91

side

0.90

country

0.87

door

0.85

most

0.85

Activations Density 0.042%

references to locations inside a vehicle, specifically focusing on the back seat

references to the locations of individuals within vehicles

No Comments

No Known Activations