INDEX

Explanations

the word "och"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "och" 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

REC

-0.67

 FedEx

-0.61

Bird

-0.58

ERC

-0.57

GMT

-0.57

ŃĶ

-0.57

 Archangel

-0.56

 variance

-0.56

 permission

-0.56

DMV

-0.56

POSITIVE LOGITS

rane

1.07

orus

1.05

lear

1.00

schild

1.00

ocolate

0.96

owa

0.93

rome

0.92

och

0.90

spr

0.88

ambers

0.88

Activations Density 0.013%

the word "och"

occurrences of the term "och" in various contexts

No Comments

No Known Activations