INDEX

Explanations

the word "rig" with varying degrees of intensity

oai_token-act-pair · gpt-3.5-turbo

occurrences and variations of the word "rig."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

Dot

-0.81

 Hots

-0.74

 Peninsula

-0.74

 Upton

-0.68

kHz

-0.68

Cf

-0.65

 Franch

-0.65

 Apostle

-0.63

 Malays

-0.63

 Pesh

-0.63

POSITIVE LOGITS

orously

1.67

idity

1.12

rig

1.11

gers

1.04

rig

1.00

rador

0.95

aton

0.91

 rigs

0.89

gins

0.89

yrinth

0.88

Activations Density 0.011%

the word "rig" with varying degrees of intensity

occurrences and variations of the word "rig."

No Comments

No Known Activations