INDEX

Explanations

phrases introducing a reason or explanation

oai_token-act-pair · gpt-3.5-turbo

explanations or justifications for statements

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

 defence

-0.64

 fresh

-0.63

 demol

-0.63

 rank

-0.62

 advanced

-0.62

 familiar

-0.61

 ranks

-0.61

 skirm

-0.60

 backfield

-0.60

far

-0.60

POSITIVE LOGITS

 Because

3.09

Because

2.75

because

1.96

ecause

1.81

 Therefore

1.68

cause

1.67

Cause

1.63

 Whereas

1.58

 Although

1.57

 because

1.55

Activations Density 0.024%

phrases introducing a reason or explanation

explanations or justifications for statements

No Comments

No Known Activations