INDEX

Explanations

names of people or characters

oai_token-act-pair · gpt-3.5-turbo

names and identifiers, likely related to people or organizations

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

 Vaugh

-0.70

 nodd

-0.70

ãĥ¼ãĥĨãĤ£

-0.67

akespe

-0.65

renheit

-0.65

 colle

-0.64

ilaterally

-0.64

xxxx

-0.62

 prest

-0.62

ãĥ¼ãĥĨ

-0.62

POSITIVE LOGITS

 Shin

0.75

 Mysteries

0.58

 Release

0.56

 Shuttle

0.55

Hos

0.54

 Savior

0.54

Ĥª

0.54

 Ahmad

0.53

 Advocate

0.53

 Chin

0.53

Activations Density 0.531%

names of people or characters

names and identifiers, likely related to people or organizations

No Comments

No Known Activations