INDEX

Explanations

phrases related to news broadcasts

oai_token-act-pair · gpt-3.5-turbo

phrases or symbols indicating a particular edition or reference point in time

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 sacrific

-0.80

 unintended

-0.74

onite

-0.69

 disadvant

-0.69

 intercepted

-0.68

 triangles

-0.67

 levers

-0.66

 mathemat

-0.66

 phosphorus

-0.63

 smugglers

-0.63

POSITIVE LOGITS

ï¸ı

1.35

ï¸

0.99

ship

0.88

âĸł

0.87

iversary

0.87

VERTISEMENT

0.85

eric

0.78

cation

0.78

early

0.76

forward

0.76

Activations Density 0.308%

phrases related to news broadcasts

phrases or symbols indicating a particular edition or reference point in time

No Comments

No Known Activations