INDEX

Explanations

names and words related to a person named Raj

oai_token-act-pair · gpt-3.5-turbo

references to a specific individual named Raj

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

ãĥĩãĤ£

-0.73

ãĤ¦

-0.73

POS

-0.71

 Yankee

-0.68

éļ

-0.68

âĻ¦

-0.67

tein

-0.67

çĦ

-0.64

ä¹ĭ

-0.62

ãĤ¤

-0.62

POSITIVE LOGITS

arat

1.13

aram

0.92

endra

0.90

alla

0.87

aji

0.86

Raj

0.86

ournal

0.86

uling

0.85

cliffe

0.85

ataka

0.83

Activations Density 0.015%

names and words related to a person named Raj

references to a specific individual named Raj

No Comments

No Known Activations