INDEX

Explanations

directional cues related to geographical locations, particularly focusing on the west direction

oai_token-act-pair · gpt-3.5-turbo

references to directional locations, particularly "west" and "east."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

staking

-0.82

displayText

-0.79

brance

-0.77

inventoryQuantity

-0.76

 reluct

-0.75

aminer

-0.74

channelAvailability

-0.73

Number

-0.73

 confir

-0.73

TABLE

-0.72

POSITIVE LOGITS

ward

1.25

side

1.08

west

1.04

 west

1.01

 coast

0.94

wards

0.93

fall

0.84

 east

0.84

urst

0.83

 hemisphere

0.82

Activations Density 0.015%

directional cues related to geographical locations, particularly focusing on the west direction

references to directional locations, particularly "west" and "east."

No Comments

No Known Activations