INDEX

Explanations

mentions of specific temperature values

oai_token-act-pair · gpt-3.5-turbo

temperature-related descriptions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Dupl

-0.82

chini

-0.76

 Infrastructure

-0.71

iture

-0.70

NetMessage

-0.67

nda

-0.66

ware

-0.66

 Spending

-0.64

ilage

-0.64

cash

-0.63

POSITIVE LOGITS

 lows

1.03

 highs

1.00

 Fahrenheit

0.99

 temperatures

0.97

 colder

0.95

 Celsius

0.93

LOW

0.88

 temperature

0.87

 hotter

0.86

 warmer

0.85

Activations Density 0.178%

mentions of specific temperature values

temperature-related descriptions

No Comments

No Known Activations