INDEX

Explanations

references to technology, specifically related to the brand "EM."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the acronym "EM."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

netflix

-0.77

 skirts

-0.69

 otherwise

-0.69

 Kubrick

-0.66

 Strauss

-0.66

 Kissinger

-0.65

 Penguin

-0.63

Cab

-0.62

 Alaska

-0.60

sid

-0.60

POSITIVE LOGITS

ERG

1.25

igrants

1.08

achine

1.06

BLE

1.05

PL

1.00

EMBER

1.00

igrant

0.99

icro

0.98

PLE

0.96

ILY

0.96

Activations Density 0.005%

references to technology, specifically related to the brand "EM."

occurrences of the acronym "EM."

No Comments

No Known Activations