INDEX

Explanations

generic positive and supportive phrases

oai_token-act-pair · gpt-3.5-turbo

conjunctions and phrases that indicate a buildup or accumulation of information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

iots

-0.76

pmwiki

-0.72

enger

-0.70

ciples

-0.65

thood

-0.64

asse

-0.63

arro

-0.63

ãĤ´ãĥ³

-0.63

emn

-0.62

iot

-0.62

POSITIVE LOGITS

 deserves

1.40

 illustrates

1.16

 underscores

1.13

 demonstrates

1.13

 reinforces

1.13

 represents

1.09

 constitutes

1.07

 reflects

1.05

 signifies

1.04

 reminds

1.04

Activations Density 0.230%

generic positive and supportive phrases

conjunctions and phrases that indicate a buildup or accumulation of information

No Comments

No Known Activations