INDEX

Explanations

pronouns or possessive words indicating relationships between different entities

oai_token-act-pair · gpt-3.5-turbo

references to relationships and interactions between people

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

aneously

-0.70

stals

-0.68

ctors

-0.67

unts

-0.63

stein

-0.62

monds

-0.60

aneous

-0.60

-0.58

mons

-0.57

ships

-0.57

POSITIVE LOGITS

hip

0.97

heet

0.96

hare

0.93

etter

0.93

etting

0.91

mith

0.91

ilver

0.90

cape

0.81

peed

0.80

erver

0.79

Activations Density 0.294%

pronouns or possessive words indicating relationships between different entities

references to relationships and interactions between people

No Comments

No Known Activations