INDEX

Explanations

mentions of geographical locations, specifically seas

oai_token-act-pair · gpt-3.5-turbo

mentions of various seas

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

cussion

-0.81

Interstitial

-0.79

arantine

-0.78

IBLE

-0.76

icating

-0.76

ÑĮ

-0.75

icator

-0.74

ication

-0.73

ijah

-0.72

ivalent

-0.71

POSITIVE LOGITS

Sea

0.97

 Scrolls

0.92

 Deity

0.82

plane

0.81

 Islands

0.79

 Serpent

0.78

 Ocean

0.77

 Isle

0.75

Sea

0.75

 monster

0.74

Activations Density 0.006%

mentions of geographical locations, specifically seas

mentions of various seas

No Comments

No Known Activations