INDEX

Explanations

general statements or facts described with the word "These"

oai_token-act-pair · gpt-3.5-turbo

the phrase "These" followed by numerical values indicating changes or observations

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

pless

-0.81

rium

-0.69

mur

-0.67

tops

-0.66

orate

-0.66

nah

-0.62

iness

-0.61

onement

-0.61

orial

-0.60

rogram

-0.60

POSITIVE LOGITS

 fellows

0.76

 findings

0.73

 distinctions

0.73

 Guidelines

0.70

 Costs

0.70

 Useful

0.69

 contrasts

0.69

 Changes

0.68

 arrang

0.68

ortium

0.68

Activations Density 0.019%

general statements or facts described with the word "These"

the phrase "These" followed by numerical values indicating changes or observations

No Comments

No Known Activations