INDEX

Explanations

mentions or discussions about future plans, goals, or strategies

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ãģį

-0.77

avis

-0.75

Ton

-0.73

pmwiki

-0.71

çĶ°

-0.67

Ô

-0.66

uddin

-0.65

atern

-0.65

tons

-0.64

Case

-0.64

POSITIVE LOGITS

aries

1.36

 vision

1.08

 visions

0.94

 goggles

0.87

obs

0.86

ary

0.86

eers

0.82

eering

0.82

naire

0.79

scape

0.78

Activations Density 0.033%

mentions or discussions about future plans, goals, or strategies

No Comments

No Known Activations