INDEX

Explanations

phrases related to being recognized or honored for achievements

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "awarded" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

sen

-0.81

APD

-0.72

Alz

-0.72

ben

-0.71

ierre

-0.70

Dak

-0.67

Geo

-0.66

 Anchorage

-0.65

andem

-0.61

 Stur

-0.61

POSITIVE LOGITS

 awarded

1.51

 awarding

1.31

 awards

1.01

 prizes

1.00

 award

0.92

 bestowed

0.86

 rewarded

0.81

Winner

0.78

interstitial

0.78

 granted

0.78

Activations Density 0.007%

phrases related to being recognized or honored for achievements

occurrences of the word "awarded" in various contexts

No Comments

No Known Activations