INDEX

Explanations

phrases related to similarities or patterns

oai_token-act-pair · gpt-3.5-turbo

instances of similarities or comparisons in context

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

yrights

-0.85

ãĥĳ

-0.85

acers

-0.84

brates

-0.84

hens

-0.82

eday

-0.80

mercial

-0.80

apps

-0.78

meric

-0.77

osponsors

-0.74

POSITIVE LOGITS

 predicament

1.54

 dilemma

1.43

 imbalance

1.37

 trend

1.36

 phenomenon

1.34

 situation

1.32

 tendency

1.31

 disparity

1.30

 plight

1.27

 frustration

1.23

Activations Density 0.372%

phrases related to similarities or patterns

instances of similarities or comparisons in context

No Comments

No Known Activations