INDEX

Explanations

mentions of locations in China, particularly related to technological or political events

oai_token-act-pair · gpt-3.5-turbo

proper nouns related to specific geographical locations or 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

croft

-0.83

leck

-0.74

 steroids

-0.71

oway

-0.71

 Adin

-0.71

lete

-0.69

ament

-0.68

 Keane

-0.66

crow

-0.65

miss

-0.64

POSITIVE LOGITS

zhou

1.19

jiang

1.01

hao

0.93

acters

0.88

hai

0.88

Xia

0.86

pload

0.84

hua

0.81

alan

0.81

 Jiang

0.78

Activations Density 0.035%

mentions of locations in China, particularly related to technological or political events

proper nouns related to specific geographical locations or entities

No Comments

No Known Activations