INDEX

Explanations

phrases related to tax breaks and cuts

oai_token-act-pair · gpt-3.5-turbo

references to tax-related terms and policies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Sensor

-0.76

Ul

-0.70

ãĤ®

-0.68

igans

-0.65

 Turk

-0.64

ãĤ¨ãĥ«

-0.61

ãĥ¥

-0.60

chat

-0.60

Nag

-0.60

ãģ£

-0.59

POSITIVE LOGITS

payer

1.01

 amnesty

0.91

 giveaway

0.89

 benefiting

0.88

 giveaways

0.82

riage

0.82

 Marketplace

0.80

amacare

0.77

 boon

0.76

benefit

0.76

Activations Density 0.111%

phrases related to tax breaks and cuts

references to tax-related terms and policies

No Comments

No Known Activations