INDEX

Explanations

names that end with "ns"

oai_token-act-pair · gpt-3.5-turbo

mentions of the entity "Laurens."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 bearer

-0.67

 slee

-0.65

 repetitive

-0.65

 medium

-0.64

 doub

-0.63

 concluding

-0.62

CONCLUS

-0.61

 speculative

-0.59

thumbnails

-0.59

 subdiv

-0.59

POSITIVE LOGITS

ertodd

1.35

ns

1.13

fw

1.02

ifiers

0.98

hers

0.98

earch

0.98

ync

0.95

etsk

0.94

uit

0.93

ettings

0.92

Activations Density 0.006%

names that end with "ns"

mentions of the entity "Laurens."

No Comments

No Known Activations