INDEX

Explanations

text related to instructions or guidelines

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "instructions" in various contexts

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

osphere

-0.73

bnb

-0.73

ade

-0.70

cture

-0.70

pole

-0.68

itivity

-0.67

deal

-0.66

gdala

-0.66

itton

-0.66

anthrop

-0.65

POSITIVE LOGITS

 instructions

3.88

 Instructions

2.63

 instruction

2.42

 directions

2.28

Instruct

1.92

 Instruction

1.80

 commands

1.59

 orders

1.58

 Directions

1.54

 directives

1.53

Activations Density 0.006%

text related to instructions or guidelines

occurrences of the word "instructions" in various contexts

No Comments

No Known Activations