INDEX

Explanations

phrases related to weapons and military actions

oai_token-act-pair · gpt-3.5-turbo

references to historical events or conditions related to conflict and its consequences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 eleph

-1.15

 pione

-0.99

 exting

-0.96

 conduc

-0.96

 streng

-0.94

 conclud

-0.93

 destro

-0.90

 tremend

-0.89

ortunately

-0.86

 unnecess

-0.86

POSITIVE LOGITS

under

1.05

appropriate

0.96

inar

0.92

unders

0.91

emphasis

0.90

lore

0.90

urban

0.90

unc

0.89

lance

0.86

apest

0.85

Activations Density 0.141%

phrases related to weapons and military actions

references to historical events or conditions related to conflict and its consequences

No Comments

No Known Activations