INDEX

Explanations

phrases indicating uncertainty or possibility

oai_token-act-pair · gpt-3.5-turbo

instances of uncertainty or tentative statements

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

oun

-0.75

 wagon

-0.72

 mosqu

-0.72

 rall

-0.71

 compr

-0.71

 advoc

-0.70

bil

-0.67

perty

-0.67

onga

-0.66

 withd

-0.65

POSITIVE LOGITS

 Absolutely

1.03

 Especially

0.98

But

0.97

 Because

0.96

org

0.94

 Anyway

0.94

 Unless

0.94

 Secondly

0.92

And

0.92

 That

0.91

Activations Density 0.458%

phrases indicating uncertainty or possibility

instances of uncertainty or tentative statements

No Comments

No Known Activations