INDEX

Explanations

first person pronouns ('I', 'we', 'me', etc.) followed by verbs

oai_token-act-pair · gpt-3.5-turbo

instances of personal beliefs or thoughts expressed by the speaker

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

adra

-0.90

catentry

-0.70

Production

-0.68

Footnote

-0.65

ario

-0.64

ante

-0.64

 horizont

-0.64

ascript

-0.63

restling

-0.62

iment

-0.61

POSITIVE LOGITS

 Bout

0.72

 provoking

0.68

 Chaff

0.66

cho

0.65

Goo

0.64

ovie

0.63

ould

0.62

 overboard

0.61

uchi

0.59

osaurs

0.57

Activations Density 0.184%

first person pronouns ('I', 'we', 'me', etc.) followed by verbs

instances of personal beliefs or thoughts expressed by the speaker

No Comments

No Known Activations