INDEX

Explanations

phrases related to dispute or conflict

oai_token-act-pair · gpt-3.5-turbo

phrases containing varying degrees of negativity or strong criticism

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

 Gaul

-0.75

 Rudd

-0.75

SAM

-0.73

 Polk

-0.72

Nau

-0.69

 Slug

-0.69

 Monkey

-0.67

 Doodle

-0.66

Hud

-0.65

 Filter

-0.64

POSITIVE LOGITS

extremely

1.15

responsible

1.14

expected

1.12

reci

1.10

emb

1.09

treated

1.08

very

1.07

sufficient

1.05

absolutely

1.05

really

1.05

Activations Density 0.093%

phrases related to dispute or conflict

phrases containing varying degrees of negativity or strong criticism

No Comments

No Known Activations