INDEX

Explanations

proper nouns or specific terms associated with various entities or concepts, potentially within a legal or legislative context

oai_token-act-pair · gpt-3.5-turbo

abbreviations or acronyms related to organizations or concepts

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

ãĥŁ

-0.74

 traged

-0.67

NetMessage

-0.67

 margins

-0.64

natureconservancy

-0.63

Reviewer

-0.63

Preview

-0.62

ħĭ

-0.62

ustomed

-0.61

 onions

-0.61

POSITIVE LOGITS

bag

0.72

alus

0.68

arb

0.68

hower

0.68

hyde

0.67

minus

0.67

geon

0.65

 Dill

0.65

zig

0.65

haus

0.63

Activations Density 0.124%

proper nouns or specific terms associated with various entities or concepts, potentially within a legal or legislative context

abbreviations or acronyms related to organizations or concepts

No Comments

No Known Activations