INDEX

Explanations

words related to specific processes or systems

oai_token-act-pair · gpt-3.5-turbo

terms related to processes and systems, particularly in contexts of evaluation and selection

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

ç·

-0.73

å§«

-0.67

ãĤ¤ãĥĪ

-0.66

thumbnails

-0.66

APTER

-0.64

inventoryQuantity

-0.62

æ©

-0.61

ãģĦ

-0.60

clinton

-0.60

ãĥ©ãĥ³

-0.59

POSITIVE LOGITS

 overseen

0.61

raught

0.58

elist

0.57

hops

0.57

 hose

0.56

phrase

0.56

 coupled

0.55

 extraord

0.55

 geared

0.54

ablished

0.54

Activations Density 0.421%

words related to specific processes or systems

terms related to processes and systems, particularly in contexts of evaluation and selection

No Comments

No Known Activations