INDEX

Explanations

names of people

oai_token-act-pair · gpt-3.5-turbo

proper nouns, specifically names and entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 behav

-0.83

MU

-0.77

Mi

-0.74

 Mike

-0.72

 distingu

-0.71

 Mister

-0.71

 Miss

-0.70

 mixed

-0.70

 misunder

-0.70

Dw

-0.69

POSITIVE LOGITS

anyahu

1.07

EStreamFrame

1.06

otti

1.05

arten

1.03

owitz

1.02

hester

1.00

hler

1.00

burgh

0.99

artz

0.98

berg

0.96

Activations Density 0.160%

names of people

proper nouns, specifically names and entities

No Comments

No Known Activations