INDEX

Explanations

mentions of flooding or flood-related terms

oai_token-act-pair · gpt-3.5-turbo

references to flooding and related water events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

cius

-0.75

 Lifetime

-0.74

 Assass

-0.74

CRIPTION

-0.66

Eb

-0.66

CRIP

-0.66

Ģ

-0.63

 Weasley

-0.63

ccording

-0.63

VICE

-0.62

POSITIVE LOGITS

waters

1.51

plain

1.18

water

1.07

 waters

1.03

lights

1.02

proof

0.94

 plains

0.93

storms

0.92

lighting

0.90

 inund

0.90

Activations Density 0.043%

mentions of flooding or flood-related terms

references to flooding and related water events

No Comments

No Known Activations