INDEX

Explanations

phrases indicating intention or plans to do something

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate a specific type of group or categorization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

abases

-0.76

terness

-0.73

Mub

-0.69

 underest

-0.67

abilities

-0.66

 flaws

-0.65

!/

-0.65

 amounts

-0.64

asley

-0.64

erity

-0.63

POSITIVE LOGITS

 centerpiece

0.99

 viable

0.85

 cohesive

0.82

 reality

0.80

 seamless

0.79

 permanent

0.79

 compulsory

0.78

Ħ¢

0.77

 profitable

0.77

 worthwhile

0.77

Activations Density 0.191%

phrases indicating intention or plans to do something

phrases that indicate a specific type of group or categorization

No Comments

No Known Activations