INDEX

Explanations

phrases that involve reflecting on or showing characteristics or attributes

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ciating

-0.72

corn

-0.71

icz

-0.70

 externalToEVAOnly

-0.69

queue

-0.69

worms

-0.67

 Unloaded

-0.66

phans

-0.65

 dangling

-0.62

é¾įå

-0.61

POSITIVE LOGITS

 poorly

0.95

 negatively

0.93

 reality

0.92

ively

0.90

 realities

0.90

 positively

0.88

 adversely

0.88

 badly

0.85

 sunlight

0.84

 reflections

0.80

Activations Density 0.054%

phrases that involve reflecting on or showing characteristics or attributes

No Comments

No Known Activations