INDEX

Explanations

specific mentions of built-in features or components in devices or software

oai_token-act-pair · gpt-3.5-turbo

phrases related to built-in features or components of devices

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Tart

-0.65

Swe

-0.64

Lol

-0.58

 Ryder

-0.57

 Â¯

-0.56

 diplom

-0.55

 bothered

-0.55

wil

-0.54

 Shapiro

-0.54

hus

-0.54

POSITIVE LOGITS

up

0.94

fledged

0.92

library

0.82

estate

0.80

essential

0.80

ups

0.80

condition

0.79

in

0.78

fill

0.76

around

0.76

Activations Density 0.035%

specific mentions of built-in features or components in devices or software

phrases related to built-in features or components of devices

No Comments

No Known Activations