INDEX

Explanations

the word "don't" or similar contractions

oai_token-act-pair · gpt-3.5-turbo

the word "don't" and its variations

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

 exha

-0.67

pus

-0.66

 watered

-0.63

 Axis

-0.60

phal

-0.59

 Mandatory

-0.58

 labour

-0.57

 dimensional

-0.57

 billing

-0.55

 misdem

-0.55

POSITIVE LOGITS

't

1.83

ning

1.20

ovan

1.16

ALD

1.13

uts

1.12

ned

1.11

keys

1.10

etsk

1.07

ations

1.05

Â´

1.02

Activations Density 0.063%

the word "don't" or similar contractions

the word "don't" and its variations

No Comments

No Known Activations