INDEX

Explanations

phrases or clauses pointing out something intriguing or significant

oai_token-act-pair · gpt-3.5-turbo

statements that express opinions or observations about various subjects

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

ishers

-0.75

Yor

-0.69

chairs

-0.68

ped

-0.67

ahs

-0.66

 buffs

-0.65

lass

-0.65

 whis

-0.64

ught

-0.64

 canoe

-0.64

POSITIVE LOGITS

Reviewer

0.84

KER

0.81

 Solitaire

0.77

FFER

0.75

GRE

0.74

 ALWAYS

0.73

PLE

0.73

VALUE

0.71

Ĥİ

0.69

ADRA

0.69

Activations Density 0.114%

phrases or clauses pointing out something intriguing or significant

statements that express opinions or observations about various subjects

No Comments

No Known Activations