INDEX

Explanations

word pairs where the first word indicates a source of information and the second word conveys a statement or action

oai_token-act-pair · gpt-3.5-turbo

the phrase "according to," indicating references or attributions in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

otin

-0.68

 heights

-0.66

obyl

-0.64

igmat

-0.64

eg

-0.63

hower

-0.63

inant

-0.62

cession

-0.61

iris

-0.58

oppable

-0.58

POSITIVE LOGITS

 sources

0.78

 ANGEL

0.76

TMZ

0.67

CNS

0.67

 Reuters

0.64

 Alto

0.64

WP

0.64

 Battalion

0.61

axter

0.61

 Radar

0.60

Activations Density 0.054%

word pairs where the first word indicates a source of information and the second word conveys a statement or action

the phrase "according to," indicating references or attributions in the text

No Comments

No Known Activations