INDEX

Explanations

verbs referring to following steps or directions

oai_token-act-pair · gpt-3.5-turbo

references to instructions or guidance

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

olson

-0.72

 ILCS

-0.71

xious

-0.69

itch

-0.69

 martyr

-0.62

rowd

-0.59

riv

-0.59

itton

-0.59

olls

-0.59

vill

-0.58

POSITIVE LOGITS

 instructions

0.98

 booklet

0.98

 manuals

0.96

eering

0.95

Instruct

0.86

auri

0.85

 instruction

0.85

 Instructions

0.83

 confir

0.83

 directions

0.82

Activations Density 0.035%

verbs referring to following steps or directions

references to instructions or guidance

No Comments

No Known Activations