INDEX

Explanations

words related to amplification and enhancement

oai_token-act-pair · gpt-3.5-turbo

terms related to amplification and increasing intensity

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

hov

-0.79

sm

-0.76

keye

-0.74

nee

-0.74

neys

-0.71

Wal

-0.71

pper

-0.70

hire

-0.70

ERAL

-0.70

ALE

-0.69

POSITIVE LOGITS

 ampl

1.50

 amplification

1.43

 amplify

1.30

 amplified

1.24

 amplifier

1.14

 Ampl

1.00

 cords

0.93

ifiers

0.91

 amps

0.86

 louder

0.86

Activations Density 0.010%

words related to amplification and enhancement

terms related to amplification and increasing intensity

No Comments

No Known Activations