INDEX

Explanations

cave-related terms

oai_token-act-pair · gpt-3.5-turbo

repetitions of the word "cave."

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

•Caves

No Comments

Negative Logits

pring

-0.83

abeth

-0.73

icious

-0.70

 ACTIONS

-0.69

nesota

-0.68

imated

-0.67

 ILCS

-0.65

ICS

-0.64

uckles

-0.64

oppable

-0.64

POSITIVE LOGITS

 cave

1.06

 caves

0.99

 Dwell

0.91

 paintings

0.86

 entrances

0.83

yrinth

0.76

 canyon

0.76

 diving

0.72

 lions

0.71

 walls

0.71

Activations Density 0.025%

cave-related terms

repetitions of the word "cave."

No Comments

No Known Activations