INDEX

Explanations

the word "Only" followed by a number

oai_token-act-pair · gpt-3.5-turbo

the word "Only" indicating a limitation or exclusivity in statements

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

 charism

-0.80

insula

-0.72

multipl

-0.64

senal

-0.63

 shenan

-0.63

arted

-0.62

dict

-0.61

sem

-0.61

raught

-0.60

ung

-0.59

POSITIVE LOGITS

ices

0.79

 marginally

0.77

onso

0.76

ICES

0.72

 incidentally

0.72

 kidding

0.71

oor

0.70

 accepts

0.70

thia

0.68

phies

0.67

Activations Density 0.055%

the word "Only" followed by a number

the word "Only" indicating a limitation or exclusivity in statements

No Comments

No Known Activations