INDEX

Explanations

phrases referring to events happening in the past

oai_token-act-pair · gpt-3.5-turbo

references to individuals in positions of authority or their titles in context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

toe

-0.70

illon

-0.64

icles

-0.63

 fulfillment

-0.62

 Parables

-0.58

 triangles

-0.58

md

-0.58

 Shore

-0.57

acts

-0.57

kson

-0.57

POSITIVE LOGITS

Reviewer

0.78

 Yugoslav

0.74

 proceeded

0.72

adays

0.70

å§«

0.69

-'

0.67

 Yugoslavia

0.67

nih

0.65

 unpublished

0.63

vernment

0.63

Activations Density 0.054%

phrases referring to events happening in the past

references to individuals in positions of authority or their titles in context

No Comments

No Known Activations