INDEX

Explanations

computer programming-related terms

oai_token-act-pair · gpt-3.5-turbo

special characters or non-standard text elements

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

 disadvant

-0.69

 psychiat

-0.67

 contrace

-0.65

 conclud

-0.64

 destro

-0.61

 undermin

-0.60

 encount

-0.57

anwhile

-0.57

 convol

-0.57

 helicop

-0.56

POSITIVE LOGITS

 âĢº

0.75

Ì

0.69

Â§

0.61

 âĢ

0.60

Ð¾

0.60

 à¤

0.59

 ãĥ

0.59

Ðµ

0.58

âĢ

0.58

à¸

0.58

Activations Density 0.450%

computer programming-related terms

special characters or non-standard text elements

No Comments

No Known Activations