INDEX

Explanations

phrases that introduce a statement or argument

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "that" in various contexts

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

aukee

-0.71

en

-0.69

backer

-0.68

gallery

-0.68

ien

-0.65

raq

-0.64

Guard

-0.63

wn

-0.62

anie

-0.62

AMY

-0.61

POSITIVE LOGITS

 they

0.75

 contradicts

0.75

'[

0.73

 justifies

0.70

 witches

0.69

"#

0.69

"[

0.69

 someday

0.67

we

0.65

 somehow

0.64

Activations Density 0.269%

phrases that introduce a statement or argument

occurrences of the word "that" in various contexts

No Comments

No Known Activations