INDEX

Explanations

specific distance measurements

oai_token-act-pair · gpt-3.5-turbo

instances of the word "about" followed by numerical quantities or distances

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

 Peaks

-0.75

chens

-0.73

rift

-0.69

hesis

-0.69

 Writ

-0.68

Write

-0.66

ses

-0.64

oly

-0.64

ãĤ¿

-0.63

BT

-0.62

POSITIVE LOGITS

 halfway

0.84

aleb

0.75

 midway

0.71

 Ambro

0.68

urb

0.67

PsyNetMessage

0.66

 calib

0.65

pload

0.65

clock

0.62

 Noon

0.62

Activations Density 0.113%

specific distance measurements

instances of the word "about" followed by numerical quantities or distances

No Comments

No Known Activations