INDEX

Explanations

words related to cages or containment

oai_token-act-pair · gpt-3.5-turbo

references to animal enclosures or cages

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Bundes

-0.73

ACTED

-0.72

olitan

-0.67

ibel

-0.64

 lender

-0.64

rians

-0.64

merce

-0.61

isance

-0.60

 Nare

-0.59

IELD

-0.59

POSITIVE LOGITS

 cage

1.08

 cages

1.00

pit

0.92

door

0.92

 Cage

0.90

washer

0.87

mong

0.86

Frames

0.77

bones

0.75

mares

0.75

Activations Density 0.016%

words related to cages or containment

references to animal enclosures or cages

No Comments

No Known Activations