INDEX

Explanations

random strings of characters that may not have a clear pattern or meaning

oai_token-act-pair · gpt-3.5-turbo

symbols and characters related to different languages and scripts

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

engers

-0.81

 Seah

-0.73

zees

-0.73

itudinal

-0.72

mares

-0.71

 Hole

-0.69

zee

-0.69

iage

-0.68

 Bengal

-0.67

geries

-0.67

POSITIVE LOGITS

å¥³

1.17

ption

1.14

âĸĳâĸĳ

1.05

entric

0.95

çĶŁ

0.90

LECT

0.89

ptive

0.88

âĸĳ

0.87

éĹ

0.87

ILY

0.86

Activations Density 0.015%

random strings of characters that may not have a clear pattern or meaning

symbols and characters related to different languages and scripts

No Comments

No Known Activations