INDEX

Explanations

phrases related to expressing opinions or advocating for a cause

oai_token-act-pair · gpt-3.5-turbo

terms related to expressing opinions and speaking out

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 skelet

-0.80

 Kear

-0.77

onut

-0.75

etheless

-0.74

opes

-0.69

esity

-0.68

 Lust

-0.68

amily

-0.67

raphics

-0.65

Ã£

-0.64

POSITIVE LOGITS

 louder

0.93

holder

0.90

 voices

0.88

 voice

0.84

voice

0.81

 Voice

0.76

 bearer

0.75

 loudly

0.73

 grievances

0.71

 trumpet

0.70

Activations Density 0.019%

phrases related to expressing opinions or advocating for a cause

terms related to expressing opinions and speaking out

No Comments

No Known Activations