INDEX

Explanations

names or terms related to a specific person or entity

oai_token-act-pair · gpt-3.5-turbo

the presence of names or references related to specific individuals, particularly those with the suffix "hr."

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

BALL

-0.75

cens

-0.71

 Hots

-0.70

eers

-0.70

 Grizzlies

-0.69

yip

-0.68

-0.65

 Samoa

-0.64

 Tasman

-0.64

 Wrestling

-0.64

POSITIVE LOGITS

onds

0.97

acking

0.93

acht

0.93

lich

0.93

agan

0.92

uty

0.90

mann

0.87

anging

0.87

anged

0.87

ackers

0.86

Activations Density 0.013%

names or terms related to a specific person or entity

the presence of names or references related to specific individuals, particularly those with the suffix "hr."

No Comments

No Known Activations