INDEX

Explanations

phrases related to geographic locations

oai_token-act-pair · gpt-3.5-turbo

specific names or titles, possibly related to people, places, or significant entities

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

ccording

-0.47

untled

-0.46

 hindsight

-0.45

 deduct

-0.43

":[{"

-0.42

 tremend

-0.42

 contradictions

-0.42

 Kafka

-0.42

 redundancy

-0.41

inarily

-0.41

POSITIVE LOGITS

ensis

0.67

Jr

0.64

III

0.56

eus

0.55

hover

0.54

.).

0.51

 âĵĺ

0.50

II

0.49

;;;;;;;;;;;;

0.46

ĪĴ

0.46

Activations Density 1.175%

phrases related to geographic locations

specific names or titles, possibly related to people, places, or significant entities

No Comments

No Known Activations