INDEX

Explanations

words related to strong negative emotions, particularly anger

oai_token-act-pair · gpt-3.5-turbo

expressions of anger or frustration

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

•layer 0 - anger

No Comments

Negative Logits

 Slot

-0.75

 Specifications

-0.71

Sec

-0.70

 Provided

-0.69

Craft

-0.69

 Protocol

-0.69

 Schedule

-0.68

Cert

-0.67

Â®

-0.67

 Surviv

-0.66

POSITIVE LOGITS

 angry

3.24

 enraged

2.34

 furious

2.24

 angered

2.08

 pissed

1.95

 outraged

1.94

 anger

1.94

 unhappy

1.87

 Angry

1.85

 irritated

1.84

Activations Density 0.028%

words related to strong negative emotions, particularly anger

expressions of anger or frustration

No Comments

No Known Activations