INDEX

Explanations

specific numeric patterns related to technical data or coding

oai_token-act-pair · gpt-3.5-turbo

numerical values and percentages in the text

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

Pwr

-0.64

ardless

-0.62

doors

-0.61

folios

-0.61

 Clancy

-0.60

redit

-0.60

balance

-0.60

ifully

-0.60

rooms

-0.59

raviolet

-0.58

POSITIVE LOGITS

0.88

edly

0.75

th

0.74

âĸĳâĸĳ

0.72

rd

0.71

0.70

mm

0.69

0.68

umbo

0.67

lers

0.66

Activations Density 0.028%

specific numeric patterns related to technical data or coding

numerical values and percentages in the text

No Comments

No Known Activations