INDEX

Explanations

terms related to reflecting or portraying something

oai_token-act-pair · gpt-3.5-turbo

phrases that describe reflection or representation of ideas or values

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Clover

-0.64

eas

-0.63

 Anim

-0.60

 possession

-0.58

Pul

-0.57

erto

-0.56

Eas

-0.56

ħ

-0.56

holes

-0.55

 Portug

-0.55

POSITIVE LOGITS

 reflects

3.25

 reflect

2.64

 reflected

2.48

 reflecting

2.47

reflect

2.05

 reflective

1.87

 Reflect

1.86

 reflection

1.68

 mirrors

1.45

 reflections

1.36

Activations Density 0.024%

terms related to reflecting or portraying something

phrases that describe reflection or representation of ideas or values

No Comments

No Known Activations