INDEX

Explanations

mathematical or technical terms in a structured format

oai_token-act-pair · gpt-3.5-turbo

terms related to academic or technical definitions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Wichita

-0.73

kefeller

-0.73

 Cortex

-0.68

HQ

-0.67

PCB

-0.66

 Blueprint

-0.66

KC

-0.66

BCC

-0.65

 Harley

-0.64

CDC

-0.64

POSITIVE LOGITS

 noun

1.61

tymology

1.47

 verbs

1.42

 adjective

1.37

 verb

1.28

 slang

1.20

 pronouns

1.18

verb

1.16

 adject

1.15

Definition

1.13

Activations Density 0.226%

mathematical or technical terms in a structured format

terms related to academic or technical definitions

No Comments

No Known Activations