INDEX

Explanations

phrases related to dialogues or conversations

oai_token-act-pair · gpt-3.5-turbo

dialogue segments or conversational exchanges

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 tremend

-0.78

ometimes

-0.70

 undermin

-0.69

 Skydragon

-0.68

omething

-0.62

everal

-0.61

ounter

-0.60

senal

-0.58

quartered

-0.58

aditional

-0.57

POSITIVE LOGITS

âĶĢâĶĢâĶĢâĶĢ

0.68

↵

0.66

 ante

0.64

Tee

0.61

 Adinida

0.59

à©

0.58

ESH

0.57

edia

0.57

 Deluxe

0.56

ONSORED

0.55

Activations Density 0.114%

phrases related to dialogues or conversations

dialogue segments or conversational exchanges

No Comments

No Known Activations