INDEX

Explanations

the word "these" followed by a noun

oai_token-act-pair · gpt-3.5-turbo

references to a recurring subject or object within the text

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

achus

-0.74

ppe

-0.73

iness

-0.73

hood

-0.73

onis

-0.70

heit

-0.69

Da

-0.67

zzle

-0.67

john

-0.67

rix

-0.66

POSITIVE LOGITS

 kinds

1.01

 sorts

0.90

 fellows

0.83

 particular

0.81

 proceedings

0.81

 types

0.77

 newfound

0.77

 conduc

0.75

 findings

0.75

 developments

0.74

Activations Density 0.083%

the word "these" followed by a noun

references to a recurring subject or object within the text

No Comments

No Known Activations