INDEX

Explanations

words and phrases related to loading and unloading

oai_token-act-pair · gpt-3.5-turbo

terms related to loading and unloading processes

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

 Schwarz

-0.67

 sketch

-0.65

 Herbert

-0.63

rib

-0.62

©¶æ

-0.62

Hib

-0.60

 Square

-0.60

 scratch

-0.59

 Ralph

-0.59

 Monteneg

-0.58

POSITIVE LOGITS

loads

1.47

load

1.41

loading

1.17

ocalypse

1.11

loaded

1.04

loader

0.91

duty

0.90

 loads

0.89

arded

0.88

Load

0.87

Activations Density 0.006%

words and phrases related to loading and unloading

terms related to loading and unloading processes

No Comments

No Known Activations