INDEX

Explanations

questions or statements starting with "Did" or "Does"

oai_token-act-pair · gpt-3.5-turbo

questioning phrases that explore events or actions in a speculative manner

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

-0.82

ãģ«

-0.74

ãģ§

-0.73

ãĤĴ

-0.72

natureconservancy

-0.72

Ñģ

-0.69

å§«

-0.68

ulates

-0.66

Materials

-0.65

otics

-0.65

POSITIVE LOGITS

olated

0.84

olate

0.83

nt

0.77

zens

0.76

fred

0.74

berra

0.72

aston

0.66

lahoma

0.65

 anyone

0.64

Activations Density 0.072%

questions or statements starting with "Did" or "Does"

questioning phrases that explore events or actions in a speculative manner

No Comments

No Known Activations