INDEX

Explanations

words related to specific technologies or entities, specifically those with the "MT" prefix

oai_token-act-pair · gpt-3.5-turbo

mentions of the abbreviation "MT" in various contexts

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

hold

-0.81

holes

-0.69

Boe

-0.64

 Argon

-0.63

 Hitman

-0.62

 Cald

-0.61

visible

-0.61

 Horowitz

-0.59

critical

-0.59

 Hurricanes

-0.58

POSITIVE LOGITS

MT

1.03

TL

0.87

orsi

0.84

ength

0.83

ypes

0.82

TON

0.81

hed

0.80

ullah

0.79

ument

0.78

ired

0.78

Activations Density 0.010%

words related to specific technologies or entities, specifically those with the "MT" prefix

mentions of the abbreviation "MT" in various contexts

No Comments

No Known Activations