INDEX

Explanations

phrases related to charitable organizations, specifically the Red Cross

oai_token-act-pair · gpt-3.5-turbo

mentions of humanitarian organizations and their related activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

à¸

-0.74

Town

-0.69

 wedge

-0.65

 misdem

-0.64

Wiz

-0.64

 combinations

-0.64

 Hoover

-0.63

HAR

-0.61

é¾

-0.61

THC

-0.60

POSITIVE LOGITS

earch

0.76

agh

0.71

ocobo

0.71

 helicop

0.70

FORE

0.70

uffle

0.69

 Airways

0.69

oire

0.68

enic

0.68

orean

0.68

Activations Density 0.061%

phrases related to charitable organizations, specifically the Red Cross

mentions of humanitarian organizations and their related activities

No Comments

No Known Activations