INDEX

Explanations

mentions of launching activities or campaigns, often related to political or social causes

oai_token-act-pair · gpt-3.5-turbo

instances of "launch" related to various types of initiatives or actions

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

é¾įå

-0.74

Ö¼

-0.66

ÄŁ

-0.66

oxide

-0.64

mean

-0.64

worn

-0.63

 Egyptians

-0.62

 matters

-0.61

 behold

-0.60

ACTED

-0.59

POSITIVE LOGITS

BALL

0.86

 countdown

0.85

wagen

0.77

hement

0.73

0.72

Interstitial

0.72

Downloadha

0.70

launch

0.70

interstitial

0.70

REAM

0.70

Activations Density 0.156%

mentions of launching activities or campaigns, often related to political or social causes

instances of "launch" related to various types of initiatives or actions

No Comments

No Known Activations