INDEX

Explanations

the word "om" often followed by a number

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of specific 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 @ 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

 Slate

-0.68

lust

-0.64

Lago

-0.64

ãģį

-0.63

AVG

-0.62

 Townsend

-0.61

 Staples

-0.60

 otherwise

-0.59

 Chargers

-0.58

 royalties

-0.58

POSITIVE LOGITS

puter

1.35

ortal

1.19

essage

1.15

icro

1.12

antle

1.11

selves

1.11

ovie

1.10

obile

1.10

achine

1.06

useum

1.06

Activations Density 0.053%

the word "om" often followed by a number

repeated mentions of specific locations or entities

No Comments

No Known Activations