INDEX

Explanations

numeric patterns related to steps, parts, or sequences

oai_token-act-pair · gpt-3.5-turbo

references to article parts and sections

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

rely

-0.71

atan

-0.70

etimes

-0.70

href

-0.67

netflix

-0.66

pect

-0.65

eworld

-0.64

ually

-0.64

istance

-0.64

 minded

-0.64

POSITIVE LOGITS

 onwards

0.79

ħĭ

0.72

ļéĨĴ

0.71

ĨĴ

0.69

 Redux

0.68

 Prototype

0.67

 VIDEOS

0.65

ITIES

0.65

 onward

0.65

ECA

0.64

Activations Density 0.060%

numeric patterns related to steps, parts, or sequences

references to article parts and sections

No Comments

No Known Activations