INDEX

Explanations

phrases related to commentary, often citing sources or authorities

oai_token-act-pair · gpt-3.5-turbo

instances of closing brackets

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ignt

-0.67

 fronts

-0.64

 benches

-0.64

 asteroids

-0.62

vier

-0.61

 belts

-0.58

Ñĭ

-0.57

izons

-0.57

Ń·

-0.57

 roof

-0.57

POSITIVE LOGITS

ccording

0.98

eely

0.78

eenth

0.73

...]

0.69

ordinate

0.68

VEN

0.68

apeake

0.66

egal

0.66

eous

0.64

gypt

0.62

Activations Density 0.033%

phrases related to commentary, often citing sources or authorities

instances of closing brackets

No Comments

No Known Activations