INDEX

Explanations

mentions of fire-related events, such as flames, smoke, and blood

oai_token-act-pair · gpt-3.5-turbo

references to fire, smoke, and injuries in distressing contexts

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

Circ

-0.73

Birth

-0.70

 Canad

-0.70

 Idol

-0.69

 Childhood

-0.69

 virtues

-0.68

IGN

-0.67

Chance

-0.66

 Campus

-0.65

Gene

-0.65

POSITIVE LOGITS

 haze

0.90

 engulf

0.89

 engulfed

0.88

 flooded

0.87

 overwhelmed

0.86

 clinging

0.84

 flakes

0.83

fog

0.82

 overtake

0.81

 crawled

0.81

Activations Density 0.174%

mentions of fire-related events, such as flames, smoke, and blood

references to fire, smoke, and injuries in distressing contexts

No Comments

No Known Activations