INDEX

Explanations

mentions of heavy snowfall

oai_token-act-pair · gpt-3.5-turbo

references to snowfall or snow-related conditions

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

ENTION

-0.82

aird

-0.74

igious

-0.73

aughs

-0.73

ISTER

-0.71

 plur

-0.71

axter

-0.70

ister

-0.69

ysis

-0.67

izabeth

-0.66

POSITIVE LOGITS

flake

1.57

fall

1.00

storms

0.99

 Leopard

0.98

boarding

0.96

tro

0.93

storm

0.93

mobile

0.89

balls

0.85

bank

0.85

Activations Density 0.014%

mentions of heavy snowfall

references to snowfall or snow-related conditions

No Comments

No Known Activations