INDEX

Explanations

instances where someone is describing a situation or action by mentioning a topic and pointing out another topic with a high level of importance

oai_token-act-pair · gpt-3.5-turbo

instances of the word "which" used to introduce relative clauses or explanations

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

apult

-0.65

ctor

-0.65

CT

-0.64

HQ

-0.62

ÂłÂłÂłÂłÂłÂłÂłÂł

-0.61

Tracker

-0.61

rolet

-0.61

Problem

-0.60

et

-0.60

redit

-0.60

POSITIVE LOGITS

soever

0.95

 guts

0.77

upon

0.73

 case

0.72

adoes

0.71

abama

0.70

ĸļ

0.67

xual

0.67

 dstg

0.63

andom

0.62

Activations Density 0.035%

instances where someone is describing a situation or action by mentioning a topic and pointing out another topic with a high level of importance

instances of the word "which" used to introduce relative clauses or explanations

No Comments

No Known Activations