INDEX

Explanations

proper nouns, specifically names of people and organizations

oai_token-act-pair · gpt-3.5-turbo

commas followed by proper nouns or names

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

Match

-0.64

hift

-0.62

ãĥ¥

-0.58

 Rebirth

-0.58

 Customers

-0.57

 gaps

-0.55

 Refresh

-0.54

Cause

-0.53

 Compatibility

-0.53

 Locations

-0.53

POSITIVE LOGITS

who

0.77

 assassinated

0.76

who

0.75

 whom

0.74

whose

0.71

 Philippe

0.71

 Abdullah

0.70

 whose

0.70

responsible

0.69

told

0.69

Activations Density 0.746%

proper nouns, specifically names of people and organizations

commas followed by proper nouns or names

No Comments

No Known Activations