INDEX

Explanations

proposals or actions mentioned in legislative contexts

oai_token-act-pair · gpt-3.5-turbo

conditional phrases or expressions of intent

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

è¦ļéĨĴ

-0.82

soType

-0.71

 Sandwich

-0.70

 Ends

-0.67

GGGGGGGG

-0.67

iques

-0.65

 adolesc

-0.65

icating

-0.64

 Hits

-0.64

ourke

-0.63

POSITIVE LOGITS

 receive

1.18

be

1.12

 qualify

1.07

 need

1.05

 lose

1.04

owe

1.03

 automatically

1.03

 incur

1.00

 forfeit

0.98

 retain

0.95

Activations Density 0.162%

proposals or actions mentioned in legislative contexts

conditional phrases or expressions of intent

No Comments

No Known Activations