INDEX

Explanations

mentions or discussions related to quantities or quantities of something

oai_token-act-pair · gpt-3.5-turbo

negations and instances of absence or lack

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

 Dwell

-0.69

duc

-0.62

 Says

-0.60

aceae

-0.60

habi

-0.59

 Turtles

-0.57

allery

-0.56

 Hedge

-0.55

say

-0.55

oding

-0.54

POSITIVE LOGITS

ĪĴ

0.67

ŃĶ

0.63

 thrott

0.63

osaurs

0.62

amera

0.62

"@

0.61

 secretary

0.60

ctica

0.60

èĪ

0.58

"[

0.58

Activations Density 0.305%

mentions or discussions related to quantities or quantities of something

negations and instances of absence or lack

No Comments

No Known Activations